Discrimination and psychological distress: Does Whiteness matter for ...

Social Science & Medicine 75 (2012) 2116e2123

Contents lists available at SciVerse ScienceDirect

Social Science & Medicine journal homepage: www.elsevier.com/locate/socscimed

Discrimination and psychological distress: Does Whiteness matter for Arab Americans? Sawsan Abdulrahim a, *, Sherman A. James b, Rouham Yamout a, Wayne Baker c a

Faculty of Health Sciences, American University of Beirut, Beirut, Lebanon Sanford School of Public Policy, Duke University, United States c Department of Sociology, University of Michigan, United States b

a r t i c l e i n f o

a b s t r a c t

Article history: Available online 8 August 2012

The white racial category in the U.S. encompasses persons who have Arab ancestry. Arab Americans, however, have always occupied a precarious position in relationship to Whiteness. This study examined differences in reporting racial/ethnic discrimination among Arab Americans. It also investigated whether and how the association between discrimination and psychological distress varies by characteristics that capture an Arab American’s proximity to/distance from Whiteness. We used data from the Detroit Arab American Study (2003; n ¼ 1016), which includes measures of discrimination and the Kessler-10 scale of psychological distress. A series of logistic regression models were specified to test the discrimination epsychological distress association, stratified by five measures that capture Whiteness e subjective racial identification, religion, skin color, ethnic centrality, and residence in the ethnic enclave. Discrimination was more frequently reported by Muslim Arab Americans, those who racially identify as nonwhite, and who live in the ethnic enclave. Conversely, the association between discrimination and psychological distress was stronger for Christian Arab Americans, those who racially identify as white, who have dark skin color, and who live outside the ethnic enclave. Even though Arab Americans who occupy an identity location close to Whiteness are less subjected to discrimination, they are more negatively affected by it. The findings illuminate the complex pathways through which discrimination associates with psychological distress among ‘white’ immigrants. Further research on discrimination and health among Arab Americans can help unpack the white racial category and deconstruct Whiteness. Ó 2012 Elsevier Ltd. All rights reserved.

Keywords: Discrimination Psychological distress Whiteness Arab Americans United States

Introduction Discrimination has received increasing recognition as one of the main mechanisms to explain racial and ethnic inequities in health in the United States (U.S.) (Gee & Ford, 2011; Williams & Mohammed, 2009). Racial/ethnic discrimination is a chronic stressor that arouses physiological responses such as anger, frustration, and helplessness. These stress responses, in turn, affect health directly through immune, neuroendocrine, and cardiovascular mechanisms, or indirectly through psychological coping mechanisms (Clark, Anderson, Clark, & Williams, 1999). The exponential growth in the number of empirical studies on discrimination and health over the last two decades has led to the publication of exhaustive reviews on the subject (Krieger, 1999; Paradies, 2006; Williams & Mohammed, 2009; Williams, Neighbors, & Jackson,

* Corresponding author. Tel.: þ961 (0) 1 350 000; fax: þ961 (0) 1 744 470. E-mail address: [email protected] (S. Abdulrahim). 0277-9536/$ e see front matter Ó 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.socscimed.2012.07.030

2003). These reviews highlight three main themes. First, ample evidence supports the presence of a positive association between discrimination and poor health. Second, though there is empirical support for the link between discrimination and hypertension, low birth weight, and self-rated health; the strongest evidence corroborates the effect of discrimination on mental health and psychological distress. Third, the association between discrimination and poor health is conditional, whereby its strength varies by individual, group identity, and contextual influences. Research has sought to examine how sources of individual variability, such as coping style and racial centrality, can intensify or mitigate the discriminationehealth association. Personal coping responses can range from passively accepting a discriminatory situation to actively confronting it. Studies on discrimination, coping, and health have yielded mixed evidence showing that, whereas active coping generally mitigates the harmful effects of discrimination, it can also precipitate poor health outcomes (Clark et al., 1999; James, Hartnett, & Kalsbeek, 1983; Noh, Beiser, Kaspar, Hou, & Rummens, 1995; Paradies, 2006; Williams, Yu, Jackson, &

S. Abdulrahim et al. / Social Science & Medicine 75 (2012) 2116e2123

Anderson, 1997). Racial centrality, a construct which captures the strength of positive identification with one’s racial group, has been reported to play a role in buffering the discriminationehealth association. For example, Sellers and Shelton (2003) found that African Americans who view their race as a central component of their identity are less negatively affected by discrimination. Similar findings from research on other racialized groups support the importance of racial/ethnic centrality in relieving the pernicious health consequences of discrimination (Mossakowski, 2003; Walters & Simoni, 2002). In addition to coping style and racial centrality, racialization based on phenotype has received limited attention in research on discrimination and health. In many cultures, dark skin tone is negatively stereotyped and darker skinned people experience more discrimination. In the U.S., Caribbean Hispanics who are racialized as Black experience more structural discrimination (in the form of residential segregation) compared to Hispanics of mixed racial ancestry (Denton & Massey, 1989). Skin color also exerts an effect on the life chances and social mobility of immigrants in the U.S., with evidence from the New Immigrant Survey showing that lightest skin immigrants of any background report 17% higher wages compared to those who have dark skin (Hersch, 2008). Evidence supporting the mediating effect of skin color on the discriminationehealth association is scant and has primarily focused on hypertension (Klonoff & Landrine, 2000). One study by Borrell and colleagues, however, examined how discrimination and skin color interact to influence the self-rated and mental health of African Americans (Borrell, Kiefe, Williams, Diez-Roux, & GordonLarsen, 2006). Even though their findings did not reveal a skin color effect, the study nonetheless evoked the important question of whether and how this phenotypic characteristic may modify the discriminationehealth association. In the U.S., much of the conceptual knowledge on racial/ethnic discrimination and its complex association with health has been constructed around the experiences of African Americans and, increasingly so, Latino and Asian Americans (Gee, Ro, ShariffMarco, & Chae, 2009; Moradi & Risco, 2006; Viruell-Fuentes, 2007). This is understandable given that members of these groups have historically borne the brunt of racial oppression and continue to do so, as they negotiate blatant and subtle forms of discrimination in their daily lives. In qualitatively different ways, racial stratification also shapes the experiences of immigrant or ethnic groups who are officially classified as white. One such group is Arab Americans who have the option to identify as white but who hold a discursive position in relationship to U.S. racial categories. Arab Americans are highly diverse with respect to national origin, religious affiliation, and socioeconomic background. The category Arab American encompasses second-generation U.S.-born citizens who may not speak Arabic and newly arrived immigrants seeking naturalization. Further, Americans of Arab ancestry exhibit diversity in how they construct racial identities. Though the Office of Management and Budget classifies persons who have ancestry in any of the 22 Arab countries of the Middle East and North Africa as white (OMB, 1997), a significant proportion identify as non-white (Abdulrahim, 2008; de la Cruz & Brittingham, 2003). Given this diversity, examining Arab American members’ differential vulnerability to discrimination and its negative health outcomes is warranted. Racial/ethnic discrimination and Arab American health Much of the research on Arab Americans adopts an acculturation framework, which assumes that the health of immigrants improves as they integrate into a white, American mainstream. This framework has come under critique in that it casts white culture as

2117

normative and does not acknowledge the role racialization plays in impeding the social and economic integration of immigrants (Viruell-Fuentes, Miranda, & Abdulrahim, in press). Only a handful of studies have examined the impact of racial/ethnic discrimination on the health of Arab Americans. Postulating that the collective experience with discrimination in the aftermath of September 11 would have negative health consequences on Arab Americans, Lauderdale’s (2006) analysis of birth certificate data in California revealed that women of Arab ancestry who gave birth in the period immediately after the terrorist attacks experienced an elevated risk of poor birth outcomes. Similar findings could not be replicated in Michigan, where there was no difference in birth outcomes of Arab American women before and after September 11 (El-Sayed, Hadley, & Galea, 2008). In fact, Arab ancestry of mother was associated with lower risk of adverse birth outcomes compared to the general population, and mothers residing in Dearborn, a city with a high Arab American concentration, exhibited a low risk for low birth weight (El-Sayed & Galea, 2010). Though the authors postulated acculturation as a plausible mechanism, the protective effects of living in an “Arab ethnic enclave” in buffering against the adverse health effects of discrimination may also explain this observation. One of the earliest studies examining the relationship between discrimination, coping, and health among Arab Americans showed a strong relationship between self-reported discrimination and psychological distress that was modified by sense of personal control (Moradi & Hasan, 2004). The conditional association between discrimination and psychological distress was further confirmed in a recent study by Rousseau, Hassan, Moreau, and Thombs (2011) whose results showed that Muslim Arab Canadians experienced more psychological distress associated with discrimination compared to Christian Arab Canadians (Rousseau et al., 2011). Disaggregating data on Arab Americans by religious affiliation has become common practice in research studies, given the racialization of Islam since September 11 (Hagopian, 2004). Evidence highlights that Muslim Arab Americans who are assimilated into dominant society report higher levels of discrimination compared to both their less assimilated religious counterparts and to assimilated Christian Arab Americans (Awad, 2010). Whereas research has increasingly examined the buffering effects of religion and area of residence, as of yet, no study has questioned the group’s white racial status or investigated how the health of an Arab American may be differentially influenced by her/his proximity to or distance from Whiteness. Whiteness and Arab Americans An offshoot of Critical Race Theory, Whiteness studies have primarily focused on documenting the highly contested process through which European immigrant groups at the turn of the twentieth century gained white racial status (Brodkin, 1998; Ignatiev, 1995; Roediger, 2005). A historical analysis of Arab Americans’ relationship to U.S.-based racial categories provides a powerful indictment of “race” as a biological construct and exemplifies a case study in Whiteness. When the right to U.S. citizenship was premised on eligibility in the white racial category, Arabic-speaking immigrants took a similar approach as European immigrants and actively appealed for inclusion in the racial hierarchy as whites. Initially, immigrants arriving to the U.S. from modern day Syria and Lebanon based their claims to Whiteness on religious arguments and advanced that, as Christians and as the mediators of Western civilization, they are naturally eligible for inclusion in the white racial category (Gualtieri, 2009; Majaj, 2000; Samhan, 1999). However, as the number of court decisions denying them citizenship increased, members of the group began to employ new arguments that explicitly separated them from Blacks and

2118


Asians. Eventually, Arab Americans became eligible for citizenship, however, after a long process of litigating their Whiteness based on connections with Christianity and distance from other groups of color. Whiteness, or the process of constructing who is white and fit for U.S. citizenship, is not a historical phenomenon, but one that continues to shape notions of “cultural citizenship” through the racialization of immigrants (Ong, 1996). Thus, contemporary forms of racialization of Arab Americans, and Muslim Arab Americans in particular, as non-white “others” has received increasing attention in scholarly writings (Hagopian, 2004; Jamal & Naber, 2008; Shakir, 1997). Even before the terrorist attacks of September 11, 2001, numerous writings highlighted the exclusion and “othering” of Arab Americans and advanced that, in an increasingly hostile social and political environment, the white racial category no longer reflected their daily experiences (Naber, 2000; Samhan, 1999). One of the main themes in these writings is that Arab Americans are caught between two contradictory processes: racialization, through increasingly vehement stereotypes of the Arab as the quintessential “other” to a “white American mainstream,” and, simultaneously, their invisibility as part of the white racial category. Like West Indian immigrants in the U.S. who can identify racially or ethnically (Waters, 1999), Arab Americans engage in disparate racial identity formations. A qualitative study carried out in Detroit revealed heterogeneity in how members of this group interact with U.S. racial categories with some who identify as white and others who identify as non-white (Abdulrahim, 2008). How proximity to or distance from Whiteness influences Arab Americans’ experiences with discrimination and its negative health effects has yet to be explored. With few exceptions (Bhopal & Donaldson, 1998; Daniels & Schulz, 2006), the white racial category has received limited attention in the literature on racial/ethnic health disparities. It is often assumed that the white category is racially neutral and groups included in it identify ethnically. In this paper, we adopt the broad term “racial/ethnic discrimination” to capture discrimination against Arab Americans, a group whose members identify in complex ways but who are increasingly racialized and pushed outside the boundaries of Whiteness. Scholarly works in Critical Race Theory have explicated the interconnectedness between racial and ethnic identity among immigrants who become incorporated in the U.S. as either white ethnics or racialized others (Harawa & Ford, 2009). Ethnicity differs from race in that it is not an imposed category but a social identity based on attributional dimensions (i.e., culture, language, and descent) that members of a group choose to identify with. Ethnicity, however, is similar to race in that it also encompasses a relational dimension that is shaped to a large extent by racial hierarchies (Ford & Harawa, 2010). As such, though race and ethnicity in the context of the U.S. are distinctive, they intertwine through the relational dimension of ethnicity. Even though scholarly writings have pointed out the increasing racialization of Arab Americans, empirical research on discrimination and its psychological effects is limited to a few studies that do not account for the full range of diversity in subjective racial identification in the group. This study strives to integrate a Public Health Critical Race Framework, which advances that race consciousness is a prerequisite to informed research on racial/ethnic inequities in health (Ford & Airhihenbuwa, 2010), even when conducting research on the health of immigrants who are classified as white. Thus, the present study aims to examine differences in reporting racial/ethnic discrimination among Arab Americans accounting for characteristics that capture distance from or proximity to Whiteness. It also seeks to investigate how the association between discrimination and psychological distress may vary by these

characteristics. First, we hypothesize that not all Arab Americans experience discrimination at the same level but that members who exhibit distance from Whiteness (i.e., Muslims, those who have darker skin, who identify as non-white, and who exhibit ethnic centrality and reside in the ethnic enclave) will experience more discrimination. Second, we hypothesize that racial/ethnic discrimination associates with psychological distress for all Arab Americans, but that this association will vary by characteristics that determine an individual’s proximity to or distance from Whiteness. Based on the limited research findings among Arab Americans, we hypothesize that the discriminationedistress association will be stronger for Arab Americans who are distant from Whiteness (i.e., Muslims, those who have darker skin, who identify as non-white, who exhibit ethnic centrality, and who reside in the ethnic enclave). We test these hypotheses among Arab Americans in the Detroit Metropolitan Area, and carry out our analyses adjusting for sex, age, marital status, and socioeconomic variables. Methods This paper is based on the Detroit Arab American Study (DAAS), which was carried out in 2003 in the Detroit Metropolitan Area (DMA). Population estimates of Arab Americans in the DMA vary considerably depending on data source, but range from 125,000 to more than 400,000. The DMA has a unique history with respect to early Arab immigration to the U.S. and is currently home to one of the largest and most diverse Arab American communities (Detroit Arab American Study Team, 2009). Descriptive data from the DAAS show that, compared to Arab Americans nationally, the Arab American community in the DMA is slightly older, has a larger proportion of immigrants, and is socioeconomically disadvantaged (Abdulrahim & Baker, 2009). One of the distinctive features of the DMA is the presence of an ethnic enclave in the suburb of Dearborn. Only 30% of Arab Americans in the DMA, however, reside in Dearborn while the rest reside in Detroit and other suburbs (Detroit Arab American Study Team, 2009). The Detroit Arab American Study received ethical clearance from the University of Michigan IRB. The present study is based on secondary data analysis. Data The DAAS research team sampled 1016 adults who defined themselves as of Arab or Chaldean descent residing in the DMA during the period of JulyeNovember 2003. These individuals were selected through a dual sampling frame design of an area probability and a list frame. The area-probability frame component included Census tracts with 10% or more individuals who identified as having Arab or Chaldean ancestry on the 2000 Census. This frame was established through a conventional three-stage sampling: first, area segments were randomly chosen from the Census tracts; the second phase involved selecting household units from each of these area segments; and in the third phase, one adult resident was randomly selected from each household. The second frame was based upon pooling membership lists of 13 Arab and Chaldean American organizations. To avoid overlap, households located in area segments included in the first frame were removed. From this master membership list, a systematic random sample of households was drawn and one adult from each selected household was randomly chosen to participate in the study. Based on both sampling frames, 4619 households were screened and 1389 were found eligible. Data collection was carried out by trained bilingual Arab American interviewers who explained to potential participants that the study was carried out by researchers at the University of Michigan. From the 1389 eligible, 1016 adults accepted to participate in the study, yielding a response rate of 73.7%.


Variables The dependent variable, psychological distress, was assessed through the Kessler Psychological Distress Scale, K-10. The K-10 is a self-reported measure of 10 non-specific symptoms of anxiety and psychological distress. Each question was scored from 1 ¼ ‘none of the time’ to 5 ¼ ‘all of the time’. For each case, the scores of the 10 questions were summed to obtain a scale from 10 to 50. The scale was dichotomized into ‘no to low, mild, and moderate distress’ for those who scored below 30 and ‘high distress’ for those who scored 30 or above (Andrews & Slade, 2001). Racial/ethnic discrimination, the main independent variable, was assessed through five questions on whether the respondent or anyone in his/her household experienced, during the two years preceding the study, any of the following due to race, ethnicity, or religion: 1) verbal insults or abuse; 2) threatening words or gestures; 3) physical attack; 4) vandalism or destruction of property; and 5) loss of employment. A ‘no’ answer was coded as 0 and a ‘yes’ answer was coded as 1. Responses on all five questions were summed for each case and, in bivariate and multivariable analyses, the variable was dichotomized into 0 ¼ ‘no discrimination’ versus 1 ¼ ‘at least one type of discrimination’. Based on our review of the literature on discrimination and health, and scholarly writings on Arab Americans, we included five variables available in the DAAS questionnaire that capture what we label “proximity to/distance from Whiteness”: racial identification, Arab American ethnic centrality, religion, skin color, and residence in the ethnic enclave. We assessed racial identification using the Census race question, which included six categories: white, black or African American, American Indian or Alaska Native, Asian, Pacific Islander, and ‘other’. Less than two-thirds of participants (63%) selected the white racial category and 37% selected one of the other categories. The overwhelming majority of those who identified as non-white (90%) selected the ‘other’ race category and 6% selected Asian (Detroit Arab American Study Team, 2009). We dichotomized this variable in the analyses into ‘white’ versus ‘non-white’. Arab American Ethnic Centrality was based on an item in the questionnaire where participants were asked whether they self-identify as Arab Americans. Religious Affiliation was determined asking a question in the interview where participants selected which religious category they identified with. The overwhelming majority selected either Christian or Muslim. Skin color was assessed visually and recorded by the interviewer at the end of the interview. To enhance the validity and reliability of this measure, interviewers underwent training to classify the range of skin tones of persons of Arab descent into dark, medium, or light. The assessment was visual (no tools were used) and solely based on skin tone (without considering other phenotypic characteristics). The variable was dichotomized in the analyses into ‘dark/medium’ versus ‘light’. We created an area of residence variable based on whether the participant’s household was located in the ethnic enclave (i.e., in Dearborn, the suburb west of Detroit which has a large concentration of Arab Americans) or outside the enclave (i.e., Detroit or other suburbs). Further, we examined in bivariate analyses three variables that have been examined in immigrant health research as indicators of integration into U.S. society. First, we employed a question on place of birth in the DAAS questionnaire to create an immigrant status measure with two categories, immigrant and U.S.-born. A language preference measure was based on whether a participant completed the interview in Arabic or in English, according to her/his preference. Finally, citizenship status was assessed through a straightforward yes/no question on whether the participant is a U.S. citizen or not. In multivariate analysis, we adjusted for the following demographic and socio-economic variables: gender; age grouped into

2119

four categories; marital status categorized into ‘married’ versus ‘widowed, separated, divorced, or never married’; education dichotomized into ‘less than high school’ versus ‘high school or more’; and annual household income dichotomized into ‘less than $20,000’ versus ‘$20,000 or more’. Statistical analysis To test our first hypothesis, that some Arab Americans experience discrimination more than others, we calculated frequencies and proportions for all independent variables by racial/ethnic discrimination. Chi-square tests were employed to determine statistically significant differences. Logistic regression was employed to test our second hypothesis that the strength of the discriminationedistress association varies by characteristics that determine an Arab American’s proximity to/distance from Whiteness. We estimated the association of psychological distress with racial/ethnic discrimination stratified by each one of the Whiteness variables, after adjusting for gender, age, marital status, education, and income. For example, to test whether the association between discrimination and distress varies by racial identification, we conducted adjusted logistic regression analyses separately for those who identified as white and those who identified as non-white. Similar analyses were carried out for the four other Whiteness variables. We tested for interactions between each of the independent variables and discrimination by running a series of logistic regression models, whereby each model included the two main effect variables (i.e., discrimination and racial identification), and the interaction term. All multivariable analyses were carried out on weighted data to account for the sample demographics and the sampling design, using SPSS, version 16. Results Table 1 presents the proportion of Arab Americans in the study who reported having experienced any form of discrimination in the two years prior to the study, due to their race, ethnicity, or religion. Almost one quarter of participants (23.6%) reported having experienced verbal insults or abuse and 13.2% reported having experienced threatening words or gestures. Less than 5% reported having experienced any of the three other types of discriminatory acts such as physical attack, vandalism or destruction of property, or loss of employment. The table further shows that 7% of participants were classified as having ‘high distress’ based on their responses to the Kessler-10 scale. Findings from unweighted bivariate analyses on the association between discrimination and all other independent variables investigated confirmed our first hypothesis that not all Arab Americans report discrimination at the same level (Table 2). Men and women did not significantly differ in the proportion of reporting discrimination. However, a higher proportion of younger respondents, those who have a high school education or more, and Table 1 Proportion of Arab Americans in the DAAS who reported experiencing each of five different types of discrimination and who reported high psychological distress.

Type of discrimination Verbal insults or abuse Threatening words or gestures Physical attack Vandalism or destruction of property Loss of employment Psychological distress High distress

%

SE

23.6 13.2 1.8 4.4 3.6

1.33 1.06 0.42 0.64 0.58

7.0

0.80

2120


Table 2 Percent reporting at least one discriminatory experience by demographic, socioeconomic, dimensions of Whiteness, and immigrant-related variables.

Total Sex Male Female Age 18e29 30e39 40e59 60 and over Education