It is undeniable that the most pernicious effects of racism on health are exerted through institutional mechanisms that are difficult to measure in standard epidemiological studies. Residential racial segregation that refers to the physical separation of ethnoraces in different residential areas is one example of such an institutionalized mechanism. Historically, the dominant group in many racialized societies has restricted the areas in which nondominant groups could reside. In countries like the United States and South Africa, residential segregation has persisted long after the legal codes supporting them were repealed. Research from the United States reveals that residential segregation is a neglected but enduring legacy of racism that adversely affects the health of African-Americans in multiple ways. Residential segregation restricts access to education and employment opportunities and is a central mechanism by which racial differences in SEP have been created and reinforced. Accordingly, all indicators of SEP are strongly patterned by ethnorace, and ethnoracial differences in SEP contribute to ethnoracial disparities in health. Segregation also results in unhealthy physical and social environments for minority ethnoracial groups through either the poor quality or high cost of nutritious foods, a lack of appropriate recreational facilities, exposure to toxic environmental substances, a high level of criminal victimization, and more retail outlets for (and excessive marketing of ) alcohol and tobacco. Medical care is also of poorer quality in segregated areas, with segregation contributing to racial disparities in preventive, screening, diagnostic, treatment, and rehabilitation services. Furthermore, the concentrated poverty and cumulative disadvantage of segregated neighborhoods can result in increased exposure to chronic stressors. U.S. research indicates that due to residential segregation, the worst urban residential conditions for Whites are superior to the average residential conditions for Blacks. Several empirical studies have also found that highly segregated areas have higher rates of infant and adult mortality, as well as certain chronic and infectious diseases. In relation to health care specifically, recent research has demonstrated that despite presenting with the same clinical indications and being treated at the same healthcare institution with the same health insurance coverage, members of ethnoracial minorities are less likely to receive appropriate medical care (e.g., cardiovascular, renal, and general surgical procedures) than members of dominant groups such as Whites (see, e.g., Smedley et al., 2003). Furthermore, research has found that health providers contribute to this systemic racism by way of clinical decisions that are influenced by both explicit (Van Ryn et al., 2006) and implicit (Green et al., 2007) racist perceptions of minority ethnoracial groups. This is an emerging area of research and further work is required to determine what contribution racism makes to ethnoracial differentials in medical care.

Direct Studies Of Racism

A recent review by Paradies (2006b) demonstrated a strong association between self-reported racism and ill health after adjustment for a range of confounders. Evidence from longitudinal studies also suggests that selfreported racism precedes ill health rather than vice versa. Furthermore, the few dozen studies of racism and health that have included both black and white populations demonstrate that self-reported racism accounts for some of the Black–White disparity in health outcomes. The most consistent finding in this body of research is the association between racism and mental (ill) health such as psychological distress, depression, and anxiety. Racism also appears to be consistently associated with maladaptive behaviors such as smoking, alcohol, and substance misuse. Although racism has been associated with objectively measured physical health outcomes such as low birth weight and hypertension, the overall evidence for an association between self-reported racism and physical health is more equivocal than the concomitant association with mental health. There are a number of reasons that could explain why self-reported racism is more strongly associated with mental rather than physical health. Because most research in this area has measured health via self-report, it is possible that the association between racism and mental ill health is exaggerated due to biases arising from the self-reporting of both exposure and outcome (Paradies, 2006b). If not caused by measurement artifice, the relatively weaker association between racism and physical health may be due to a delayed effect of racism on physical health that is mediated by mental ill health. However, we are aware of no longitudinal studies that could shed light on this possibility as no such studies, to our knowledge, have examined the interactions between mental and physical health outcomes and self-reported racism.

Future Directions

The relatively stronger association between self-reported racism and mental health outcomes raises questions about the mechanisms by which racism affects health. It is currently unclear what combination of pathopsychological effects on the mind and/or direct or indirect (i.e., through other body systems) neurophysiological changes mediate the association between self-reported racism and ill health. As suggested by Harrell et al. (2003) and Mays et al. (2007), studies that employ pharmacological blocks and functional neuroscience approaches may shed light on the psychophysiology of racism. Researchers in the field of stress are now examining biomarkers specific to body systems (i.e., cardiovascular, neuroendocrine, immune) with evidence emerging that particular stressors differentially affect physiological systems. Evidence that racism is more strongly associated with diastolic, than systolic, blood pressure is an example of such specificity. Clearly, further physiological research on racism is required to uncover the biological processes through which this phenomenon affects health. Moreover, the continuing debate on whether racism is a form of stress or a construct separate from stress (with evidence that stress both mediates and moderates the relationship between racism and health), may also be resolved through such psychophysiological research. Other than preliminary evidence that a heightened sense of ethnoracial identity attenuates the adverse effects of self-reported racism on health, there has been limited research on moderation/mediation of the relationship between racism and health. Similarly, other than some evidence that the active coping is associated with better health outcomes than passive coping, little is known about the efficacy of various responses/reactions to racism in relation to health. There has also been a scarcity of research on the health effects of racism across ethnoracial groups, with most studies focusing only on African-Americans. Even among studies that do consider racism across ethnoracial groups, it is common to statistically adjust for ethnorace rather than undertake stratified analyses. More nuanced analysis will be required to understand the degree to which associations between self-reported racism and health vary across ethnoracial groups. As part of such research, the privileging aspects of racism for dominant ethnoracial groups (e.g., Whites) could also be investigated and further examination of intraracial racism may resolve the continuing conceptual debate centered on this phenomenon. Future research in this field of study is also required to characterize the multiple pathways through which institutionalized racism affects health. Furthermore, despite evidence of an association with ill health, limited attention has been focused on internalized racism. Work is needed to identify the best approach to assessing this phenomenon and to determine its impact on population health. A significant limitation in this field is that direct approaches to studying racism have relied too heavily on single-item and unvalidated measures of self-reported racism, which are of limited utility (Paradies, 2006b). There are, however, three instruments that have been commonly used to assess self-reported racism in this field: the Everyday Discrimination Scale (Williams et al., 1997), Experiences of Discrimination scale (Krieger et al., 2005) and the Schedule of Racist Events (Landrine et al., 2006). Given that these instruments have been subjected to psychometric validation and are able to assess different types of oppression across a range of ethnoracial groups, future research should (where possible) use these instruments rather than developing measures de novo. Only a minority of direct studies of racism have specified a time frame over which exposure to racism should be reported by respondents and there is ongoing debate on the appropriateness of specifying time frames for self-reported racism. Utsey and Ponterotto (1996) have argued that an unspecified time frame is appropriate given the long-lasting nature of racist experiences. However, it is unclear whether respondents report recent and/or highly salient/traumatic experiences of racism when responding to questions without a specified time frame. Also, Blank et al. (2004) have noted that the inclusion of an explicit time frame on surveys is necessary to estimate the rate of exposure and to avoid confounding time-series analyses. By measuring self-reported racism with and without specified time frames in the same study, as well as investigating diary methods that capture events as they occur (Hill et al., 2004), future research will be able to shed light on this unresolved issue. Most research in this emerging field has assessed self-reported interpersonal racism. As such, further research on internalized racism, vicarious racism (racism experienced by family or friends), systemic racism, and setting-specific racism (e.g., racism in the workplace) is warranted. Psychological research into the factors affecting perception, attribution, and reporting of racism (including the interplay between objective and subjective racism) is also required, especially studies that can uncover the factors affecting respondents’ retrospective estimate of racism exposure across a range of settings, contexts, and time frames.