Racial and socioeconomic status are among the most prominent types of intergroup relationships that reflect types of social inequality in American society. Where social inequalities exist, discrimination is a key feature of intergroup relationships (Jackman 1994). Discrimination can serve to reinforce the symbolic boundaries that separate the social groups from each other. Exposure to discrimination based on race has received the most attention in the research literature. Given the history of slavery and segregation based on law and custom, research has documented that discrimination based on race has affected a broad range of social outcomes for African Americans (Jaynes and Williams 1989). Despite the end of legally-imposed segregation and the expansion of opportunities for African Americans, recent qualitative studies document the persistence of discrimination in a broad range of social settings for blacks (Cose 1993; Essed 1991; Feagin 1991). Other research indicates that Hispanics and Asian Americans also experience discrimination (Kim and Lewis 1994; Telles and Murgia 1990). Similarly, Sigelman and Welch's (1991) review of findings from public opinion studies indicate that African Americans continue to experience discrimination in a broad range of settings in society, with the highest levels being reported in the area of housing. Nonetheless, the overall levels of discrimination reported are relatively low. In the National Study Black Americans, 11% of respondents reported that they or a family member had been treated badly because of their race in the previous month (Williams and Chung 1997). In the national Americans’ Changing Lives survey, 47% of blacks, 12% of "discriminated whites" (Polish, Italians, and Hispanics), and 7% of other whites responded affirmatively to a question about racial or ethnic discrimination (Jackson et al. 1997). Similarly, in the Commonwealth National Minority Health Survey, 7% of whites, 44% of blacks, 24% of Latinos, and 56% of Asians reported that they had been treated badly because of their race or cultural background in the previous year. In the 1995 Detroit Area Study, 68% of African Americans and 16% of whites indicated that they had been treated badly because of their race or ethnicity over their lifetime. A review of public opinion polls found that no more than 40% of blacks indicated experiencing discrimination in education, housing, jobs or equal wages, although 60% reported that they had experienced at least one of these types of discrimination (Sigelman and Welch 1991). Some research suggests that the reported levels of discrimination are importantly linked to the nature of the assessment. In particular, multiple questions that attempt to get at various components of discrimination result in higher reported levels (Bobo and Suh 1995; Sigelman and Welch 1991; Gary 1995; Krieger and Sidney 1996; Landrine and Klonoff 1996). In a sample of 4,025 African American, white, Asian, and Latino respondents, Bobo and Suh (1995) found that 59% of African Americans, 25% of whites, 22% of Asians, and 31% of Latinos reported experiencing at least one discriminatory event at their place of work.

An understanding of the impact of discrimination on health must also pay attention to the ways in which victims respond to and attempt to manage these negative experiences. Lalonde and Cameron (1994) noted that the distinction between an active versus a passive response is one of the key themes that cuts across major typologies of behavioral responses to experiences of injustice. Some limited research suggests that an active versus passive response may predict variations in the impact of reported discrimination (Krieger 1990; Harburg et al.1973). Anger is another theoretically important response pattern that has not received systematic empirical examination in the literature (Grier and Cobbs 1968).

For each measure of unfair treatment experienced, a follow-up question ascertained whether the experience had happened in the past 12 months. This scale also distinguishes past year (recent) experiences of discrimination from earlier lifetime episodes. An additional follow-up question to each endorsed discrimination item ascertained what was the main reason for unfair treatment. Respondents were allowed to select from a list that included ethnicity, gender, race, age, religion, physical appearance, sexual orientation, and income level/social class. Thus, unfair treatment attributed to race/ethnicity can be distinguished from discrimination that was attributed to other social bases.

The second type of discrimination assessed, everyday discrimination, was a measure of chronic, routine, and less overt experiences of unfair treatment (Essed 1991). Nine items captured the frequency of the following experiences in day-to-day life: being treated with less courtesy than others, less respect than others, receiving poorer service than others in restaurants or stores; people acting as if you are not smart, they are better than you, they are afraid of you, they think you are dishonest; being called names or insulted, and being threatened or harassed (Cronbach's alpha = .88).

Relationship to SES

Several recent qualitative studies of racial discrimination have focused on the black middle class (Cose 1993; Essed 1991; Feagin 1991; Feagin and Sikes 1994; Lykes 1983; Hochschild 1993). The available evidence suggests a complex pattern of association between SES and racial discrimination. Sigelman and Welch (1991) found that in studies of African Americans, education was positively associated with reports of racial discrimination but discrimination was also positively related to working class status, receipt of government assistance, and having financial difficulties. In his study of black men, Gary (1995) found that racial discrimination was positively related to both education and employment status but unrelated to income. Landrine and Klonoff (1996) also found that income was unrelated to discrimination. Bobo and Suh (1995) found a strong positive relationship between education and employment discrimination for African Americans. In multivariate analyses of the data from the Detroit Area Study, education but not income was positively related to both chronic and acute measures of discrimination (Forman et al. 1997).

Relationship to SES

Several recent qualitative studies of racial discrimination have focused on the black middle class (Cose 1993; Essed 1991; Feagin 1991; Feagin and Sikes 1994; Lykes 1983; Hochschild 1993). The available evidence suggests a complex pattern of association between SES and racial discrimination. Sigelman and Welch (1991) found that in studies of African Americans, education was positively associated with reports of racial discrimination but discrimination was also positively related to working class status, receipt of government assistance, and having financial difficulties. In his study of black men, Gary (1995) found that racial discrimination was positively related to both education and employment status but unrelated to income. Landrine and Klonoff (1996) also found that income was unrelated to discrimination. Bobo and Suh (1995) found a strong positive relationship between education and employment discrimination for African Americans. In multivariate analyses of the data from the Detroit Area Study, education but not income was positively related to both chronic and acute measures of discrimination (Forman et al. 1997).

Relationship to Health

Some limited research suggests that both the psychological and physiological correlates and consequences of discrimination are similar to those of other psychosocial stressors (Dion, Dion, and Pak 1992; Thompson 1996). Several laboratory studies have assessed the physiological and affective reactions of African Americans to mental imagery and videotaped vignettes of discriminatory behavior. They have found that such exposure to racist provocation leads to increased cardiovascular and psychological reactivity (Anderson, Myers, Pickren, and Jackson 1989; Armstead, Lawler, Gordon, Cross, and Gibbons 1989; Jones, Harrell, Morris-Prather, Thomas, and Omowale 1996; Morris-Prather et al. 1996; Sutherland and Harrell 1986). Three studies have also documented that, at least under some conditions, there is a positive relationship between reports of discrimination and blood pressure levels for blacks (James et al. 1984; Krieger 1990; Krieger and Sidney 1996) although this pattern is not uniform (Broman 1996). In addition, one study found that responses to hypothetical experiences of discrimination were related to blood pressure levels for both blacks and whites (Harburg et al. 1973).

In a series of studies in Toronto, Canada, Dion and his colleagues have documented that the experimental manipulation of discrimination in the laboratory setting is associated with increased psychological distress for Jewish undergraduate males (Dion and Earn 1975), Chinese students (Pak, Dion, and Dion 1991) and undergraduate women (Dion 1975). Two small studies of members of socially stigmatized groups in Toronto also found that self-reported discrimination was adversely related to psychological well-being. Perceptions of discrimination were positively related to psychological distress for Chinese residents (Dion et al. 1992) and reports of discrimination were inversely related to life satisfaction and personal control for gay male and lesbian adults (Birt and Dion 1987).

Several population-based studies have linked discrimination to mental health status. Thompson (1996) found that reports of discrimination were positively related to psychological distress in a probability sample of African American adults in St. Louis, Missouri. In analyses of the National Study of Black Americans, Williams and Chung (in press) found that a single item that measured the experience of racial discrimination in the previous month was positively related to psychological distress and inversely related to happiness and life satisfaction. Discrimination was also related to higher levels of chronic health problems and disability in that study. Two studies of Hispanic women have also documented a positive relationship between discrimination and psychological distress (Salgado de Snyder 1987; Amaro, Russo, and Johnson 1987). In the national Americans' Changing Lives study, perceptions of racial/ethnic discrimination were associated with physical and mental health for both blacks and whites (Jackson et al. 1997), but Krieger (1990) found that although racial discrimination was related to hypertension for black women, gender discrimination was unrelated to high blood pressure for white women.

Analyses of the 1995 Detroit Area Study documented that African Americans report higher levels of both chronic and acute experiences of unfair treatment than whites (Williams et al. 1997). Importantly, the chronic everyday measure of discrimination was more strongly linked to both psychological well-being and psychological distress than the measure of acute episodic experiences of discrimination. Instructively, discrimination affected the mental health of both African Americans and whites but the chronic measure of discrimination was more adversely related to psychological well-being of whites than of blacks (Williams et al. 1997). The Detroit study also showed that SES (especially income) accounted for most of the black-white differences on self-report indicators of physical health. However, discrimination played an incremental role in accounting for differences between the races in health status.

Limitations

Virtually all studies to date have used cross-sectional data, so that we do not know the causal direction of the observed associations. A major methodological issue in this research is the measurement of discrimination based on self-reports. That is, perceptions of discrimination could be affected by health, especially mental health status, and thereby be confounded with current levels of health. At the present time it is not clear to what extent reports of racial discrimination are the result of health status. At the same time, there is reason to believe that respondents' reports do not reflect the systematic distortion of reality. However, a series of laboratory-based experiments indicate that respondents understand the concept of discrimination as intended by the investigator and reports of personal experiences of discrimination are consistent with objective experiences (Taylor et al. 1991; Taylor et al. 1994).

Network Usage

Conclusion

The study of discrimination and health is in its infancy. Much is yet to be learned about the appropriate conceptualization and measurement of discrimination and the mechanisms and processes by which discrimination may affect health. Experiences of discrimination appear to be an important class of stressful life experiences that have been neglected by traditional approaches to the measurement of stress. Although these experiences are not limited to any segment of society they appear to be more prevalent for racial minorities and other stigmatized groups compared to the general population. Thus, it may be important to include measures of discrimination in studies of racial/ethnic minorities and in analyses of racial differences in health. Discrimination is unlikely to be a mediator of the SES-health relationship but it may moderate the association between SES and health. Discrimination may also make an incremental contribution to the health of racial minorities over and above indicators of SES.

Selected Bibliography