Death by a thousand cuts: The health implications of black respectability politics

Abstract

The authors introduce the concept of “vigilance,” capturing behaviors that reflect attempts to navigate racialized social spaces on a daily basis. Specifically, vigilant behaviors include care about appearance and language to be treated with respect, avoidance of social spaces, and psychological preparation for potential prejudice and discrimination. Furthermore, these behaviors align with those discussed in Black respectability politics debates. Using data from a population-representative sample of Black adults in Chicago, they report that vigilance is associated with poor physical and mental health indexed through chronic health conditions, depressive symptoms, and self-rated health.

Introduction

Engaging in behaviors that align with mainstream White, heterosexual, Judeo-Christian, middle class values has long been an adaptive strategy used by Black Americans in an attempt to avoid, or to at least to lessen discrimination and prejudice, ranging from the irritating to the life-threatening in a racialized and racially-stratified society such as the United States (Higginbotham 1993). Indeed, scientific and popular media narratives of Black American life are often imbued with contemplating and engaging in behaviors to validate the right to live, work, be educated, and recreate – to exist – in everyday social spaces (Du Bois 1920; Essed 1990; Feagin and Sikes 1994).

In a recent Mashable report titled “Black Armor,” millennial Black men discussed suppressing their sense of self or style to combat racial stereotypes and deflect negative attention from the police, thus using visible style as “armor” (Yi 2016). One individual interviewed discussed why, although he would like to wear casual clothes, he felt more protected in dress clothes “It’s like armor to me,” he says. “When I have a suit on I feel like all of a sudden, the world sees me differently. Cops aren’t staring, people wave back, people shake my hand, they open the door for me…”(Yi 2016).¹ Furthermore, this visible armor may actually signify a deeper psychological armor, as has been described in the scientific literature for decades. For example, in one particularly notable study on Black Americans experiences with discrimination,

One older [study] respondent spoke of having to put on her “shield” just before she leaves the house each morning…she said that for more than six decades, as she leaves her home, she has tried to be prepared for insults and discrimination in public places, even if nothing happens that day (Feagin and Sikes 1994).

The popular and scientific literatures are filled with accounts across gender and class boundaries of Black Americans adjusting their visible appearance and behavior to conform to dominant White norms, dispel Black stereotypes, and avoid prejudice and discrimination. Furthermore, these visual conformations are coupled with and signify an invisible psychological arming behavior to prepare for prejudicial and discriminatory situations.

Debates have arisen over these behavioral strategies on matters such as: their effectiveness in avoiding prejudice and discrimination; whether adapting to dominant norms results in racial denigration; their effectiveness in eliminating structural and cultural racism or advancing the well-being of the Black community; and their use to morally police one’s own race often based on class, gender, and sexuality lines (Cohen 1999; Cooper 2015; Harris 2003; Harris 2014; Houston 2015; Lebron 2015; Muhammad 2011; Nnebe 2015; Pattillo 2007; Schomburg Center 2016; White 2001). Broadly speaking, the politics of these behaviors – or the “politics of respectability” – involve the regulation of individual behavior to public presentation based on the strong desire to refute negative racial stereotypes and “…presenting one’s self as a citizen worthy of respect as defined by the dominant cultural norms and standards” (Smith 2013).²

Regardless of one’s politics regarding respectability behaviors, we propose that there is a health cost – both mentally and physically – to employing counter strategies on a regular basis. In this article, we introduce the concept of “vigilance” or “vigilant coping style,” developed in the social sciences, which captures certain behaviors that may reflect attempts to avoid individual prejudice or discrimination and navigate social spaces on a daily basis – a by-product of living in a structurally and culturally racist society (Clark, Benkert, and Flack 2006; Hicken et al. 2013; Hicken et al. 2014; Himmelstein et al. 2015). Through the lens of vigilance, we focus on certain behaviors including care about appearance and language to get good service, avoid being harassed, receive the same level of respect as White people, the avoidance of social situations and places, and preparing for potential prejudice and discrimination.

We discuss our work on vigilance and racial health inequities and then present new results on the link between vigilance and mental and physical health for Black adults, specifically. Our analysis uses data from the Chicago Community Adult Health Study (CCAHS), a cross-sectional, population-representative survey of 3,105 adults, aged 18 years and older, living in Chicago. We find that vigilant thoughts and behaviors are associated with poor self-reported health and greater risk of chronic disease and depressive symptoms. Our results suggest that engaging in behaviors that have been linked to the narrative of Black respectability politics can have important and possibly enduring health consequences for all who are forced or who choose to utilize them.

Background

Respectability Politics and Impression Management Behaviors

While interpretations of respectability politics are wide ranging, a recurrent theme involves impression management behaviors. Discussions on respectability politics either encourage or oppose the use of actions that focus on presentation of self for African Americans as a necessary or useful strategy to be safe and successful in a racist society (Goffman 1959). As Kennedy (2015) noted in his recent Harper’s essay in defense of respectability politics:

… prudent conduct and sensitivity to how we appear to others improve our chances for success in environments peppered with dangerous prejudices. It is unfortunate that safety might require such self-consciousness, and it is imperative to reform society such that self-defense of this sort is no longer needed. In the interim, however, blacks should do what they can to protect themselves against the burdens of a derogatory racial reputation that has been centuries in the making.”³

However, many argue that these behaviors are indeed not protective, as exemplified by Brittney Cooper (2015) in her Salon article entitled, “Stop poisoning the race debate: How respectability politics rears its ugly head — again: It’s not just white skeptics undermining the fight for equality. Here’s why moderate Black folks are also to blame,” Cooper directly criticized respectability politics behaviors as being ineffective in serving to improve the well-being of African Americans:

…surely you know a suit and tie won’t protect you. So we’re going to keep on marching, as you said…We will do so, because Black folks have already tested out your theory of respectability. We’ve been trying to save our lives by dressing right, talking right and never, never fucking up since about 1877. That shit has not worked.

Regardless of the multiple ways in which respectability politics and the associated behaviors are discussed, engaging in these behaviors is likely to have important implications for both mental and physical well-being. In a recent NPR Codeswitch post, by Leah Donnella (2015), she recounts a twitter conversation on the feasibility of impression management to protect Black people from police violence. This discussion ended with a tweet “Know what’s funny about this? Maybe I can do certain things to reduce my odds of getting shot. But what a burden!”

Vigilance: The Social Science of Racism and Respectability Behaviors

The social science literature has systematically documented the behaviors often discussed within the context of respectability politics (Cose 1993; Feagin 1991; Feagin and Sikes 1994; Lee 2000; Nuru-Jeter et al. 2009). A number of themes emerge from the literature regarding the nature and dimensions of these behaviors. We highlight and discuss behaviors associated with three prominent themes: (1) adapting presentation of self (including style of dress and ways of speaking) to reduce the likelihood of experiencing discrimination in social settings, (2) avoiding social situations where likelihood of discrimination may be higher, and (3) daily preparation for possible experiences with prejudice and discrimination. We then link engagement in these behaviors to deterioration in physical and mental health via the stress response pathway.

Across all levels of socioeconomic status (SES), Black Americans frequently report the need to engage in impression management (Cross 1998; Goffman 1959, 1963; Guiffrida 2003: Rosenbloom and Way 2004) and the feeling that it is necessary to be careful about their appearance to get good service, to avoid being harassed, or to be treated with the same level of respect as White people. For example, in-depth interviews with professors at Midwestern State University in 1999 found that Black professors were more likely to report over-preparing for class lectures, citing credentials, and dressing up compared to White professors, to avoid stereotyping and questioning of competency and credentials from their students (Harlow 2003; Ford 2011). One Black professor noted: “I’m always conscious how I dress on teaching days. … If I’m dressed casually on a non-teaching day and I go into someplace like the library, I frequently get the feeling that people don’t … people have absolutely no expectation that I can be a faculty member” (Ford 2011, 467).

Another form of strategic social negotiation for many African Americans may also include feeling the need to carefully monitor what they say and how they say it in an attempt to reduce the likelihood of experiencing discrimination (Aspinwall and Taylor 1997; Carter 2003; Mallett and Swim 2009; Swim and Stangor 1998). A personal narrative written by a Black researcher about his experiences in high school reflects these actions:

Because I felt that whites were constantly assessing the level of my intelligence and overall humanity, I was careful to speak with the greatest form of articulation. I made sure my statements were valid, factual, and creative. It became such an automatic response to my environment (Gadsden 2005, 265).

Blacks Americans also frequently report trying to avoid certain social situations and places where they may likely experience discrimination. In a qualitative study of Black customers in New York (NY) and Philadelphia (PA) in 1996–1997 (Lee 2000), a Black woman described her reaction of being falsely accused of theft in a department store the following way: “You have to have been a victim of those things to really understand. You get very paranoid about certain things and going to certain places because you know it’s a very hurtful thing to be accused of something you didn’t do and know that you can’t be comfortable shopping” (Lee 2000, 368). Indeed, avoidance behaviors have been identified in numerous qualitative (Brunson and Miller 2006; Lee 2000) and social psychological research settings as one possible strategy to cope with chronic experiences of discrimination and prejudice (Oyserman and Swim 2001).

Research indicates that this attention to presentation of self in preparation for or protection against prejudice and discrimination is a significant psychological burden with health consequences. We introduce a concept called “vigilance” defined as anticipatory and ruminative thoughts and behaviors involved in the preparation for discriminatory treatment and mirror behaviors that align with the presentation of self strategies encouraged by proponents of black respectability politics but likely utilized by many African Americans to engage in racially-hierarchical social spaces.

We use the term “vigilance”⁴ to characterize a psychological state that is not necessarily contingent on a prior interpersonal experience with discrimination, but is likely a by-product of attempting to function in a racialized and racially-hierarchical society and may arise through other representations of devaluation associated with membership in a stigmatized group (Goffman 1959; Harrell 2000). Vigilance may capture the experience of living in a racially-stratified social context where the stressful and health-relevant nature of racism may not only operate through the personal experience of specific discriminatory events, but also through vicarious contact, (e.g., friends’ and family members’ interpersonal experiences), contemporary and historical collective experience of inequality (e.g., racial disparities in educational opportunities and the criminal justice system, and racial segregation), racially-stratified institutionalized practices in the cultural (e.g., racially biased media portrayals) and sociopolitical environment (e.g., racially coded language in political debates), as well as through the anticipation of unfair treatment (Alexander 2010 2; Bobo 1997, 2011; Bobo and Charles 2009; Bobo and Fox 2003; Bonilla-Silva 2006; Charles 2003; Feagin 1991; Feagin and Sikes 1994; Harrell 2000; Massey 1993; Meehan 2001; Nuru-Jeter et al. 2009; Orfield, Eaton, and Harvard Project on School Desegregation 1996; Smith and Petrocelli 2001; Tyson 2011; Williams and Neighbors 2001). This may result in the need for Black people to persistently anticipate and prepare for racial discrimination as they negotiate the social spaces necessary for daily life activities (e.g., banking, shopping, and working).

That these behaviors reflect the burden of life in the US is borne out in the social science literature and in an additional dimension of vigilance – anticipation of stress. While respectability politics doesn’t explicitly include anticipatory thoughts around potential discrimination or prejudice, we offer that these invisible thoughts may actually underlie and coincide with the visible behaviors. Black Americans frequently report having to prepare for the possibility of being insulted, either overtly or covertly on a day-to-day basis. Qualitative work on discrimination has documented striking examples of the stressful toll of continually preparing oneself for racially-charged social interactions. For example, in Feagin and Sikes’ (1994) study based on in-depth interviews with middle-class Black adults about their experiences with discrimination between 1988 and 1990, one woman noted: “I feel as though most of the time I find myself being in a guarded position or somewhat on the defense. I somewhat stay prepared to be discriminated against because I never know when it’s going to happen to me” (295). Similarly, in more recent qualitative work conducted between 2004–2005 among Black women living in the San Francisco Bay area, one woman noted, “You just get tense, ‘cause you know you have to brace yourself for something stupid that they’re gonna say…with a White person, you know that some level of racism is going to hop out of their mouth…And so you have to prepare your body for that.” (Nuru-Jeter et al. 2009, 35).

The Toxicity of Vigilance

Engaging in these behaviors is not without health consequences. Indeed, many opponents of respectability politics allude to the possible mental and physical health burdens of engaging in these behaviors. In referring to Jackie Robinson, Ta-Nehisi Coates discusses in The Atlantic (2012, 2014) the “twice as good” arguments often embedded in behaviors encouraged by respectability politics (Kennedy 2015; Du Bois 1920). In “The Rage of the Privileged Class,” (2014) he states:

I was on the radio. Somebody was saying yesterday on the radio, “Well you know, Jackie Robinson did this.” And I told him, “You got to remember Jackie Robinson died young. Don’t ever forget that, every time you say that. Remember that.” You know, it wasn’t just a matter of being better. This actually costs. It costs. We should probably stop bragging about Jackie Robinson, and remember that he died young.

Science supports the notion that these chronic self-regulatory behaviors come with a health cost. We weave together evidence from different areas of the social, psychological, and biological sciences that link these behaviors to health. Then, upon this foundation, we discuss the empirical literature on the social science notion of vigilance in relation to health.

Anticipation of potential stressful situations activates the body’s primary stress response system, known as the hypothalamic pituitary adrenal axis, or HPA axis (Brosschot 2010; Brosschot, Gerin, and Thayer 2006; Brosschot, Pieper, and Thayer 2005). The HPA axis is responsible for activating physiological systems throughout the body to prepare to deal with the stressors. Activation of the stress response system occurs even in the absence of the actual occurrence of the stressful situation (Gaab et al. 2005; Smith, Ruiz, and Uchino 2000; Waugh et al. 2010). In other words, the anticipation is a critical – and healthy – component of stress biology. Nevertheless, chronic anticipatory stress, as other forms of chronic stress, results in dysfunction of the stress response system and then poor mental and physical health (McEwen 1998, 2003; Schulkin et al. 1994). Further, chronic anticipation may work in conjunction with cognitive rumination about particular stressors. Rumination, which is the repeated cognitive and affective representation of a stressor, is thought to result in prolonged biological stress responses (Brosschot 2010; Pieper and Brosschot 2005). The drawn out thought process associated with rumination can transform a discrete occurrence into a chronic situation.

Because of its ubiquitous and pervasive nature in the United States, racism, in its structural, cultural, and interpersonal forms, may provide a continual source of anticipatory and ruminative stress for Black Americans in several ways. First, and probably most studied, is personally-experienced discrimination (from everyday microaggressions to discrete major events) (Jones 1972, 1997; Sue 2010; Sue et al. 2007; Williams et al. 2012). Previous personal experiences of prejudice and discrimination may result in the anticipatory and ruminative thoughts and related behaviors captured by vigilance. Second, vicarious experiences of discrimination – either by someone within one’s social network (e.g., sister) or not (e.g., Sandra Bland) – may also result in vigilance (Graham 2015). Other drivers of vigilant thoughts and behaviors include a knowledge and understanding of structural racism (e.g., laws and practices that results in mass incarceration) and the cultural racism that results in negative stereotypes about Black Americans.

Research suggests that the anticipation of prejudice or discrimination more strongly triggers activation of the biological stress response systems compared to the anticipation of other types of stressors (Sawyer et al. 2012). For example, researchers brought two groups of Latina college students into a laboratory to apply for a hypothetical lab partner position. Part of the application process was a public speech in front of the potential lab partner. In one group, the participants were led to believe that the potential lab partner held ethnically hostile views; in the other, the participants were led to believe that the potential lab partner held views about other ethnic groups that were normative with regard to the campus. Those in former group showed greater blood pressure increases as they prepared their application speech compared to the latter group. This suggests that the anticipation of potential prejudice or discrimination results in heightened biological stress response over and above other types of stressors (i.e., public speaking).

While this anticipatory biological stress response is healthy, it is the chronic nature of the anticipatory (and ruminative) thoughts and behaviors that results in poor health. For example, research suggests that vigilance plays an important role in the well-documented racial inequalities in health. In a population-representative sample of adults in Chicago, Black adults reported greater levels of vigilance compared to White adults – and the difference in these levels explained the Black-White inequalities in poor sleep quality (Hicken et al. 2013), a health outcome that has been shown to be a critical link between the social environment and numerous chronic morbidities including cardiovascular disease and diabetes (Cappuccio et al. 2010; Knutson and Cauter 2008; Mallon, Broman, and Hetta 2002; Phillips and Mannino 2007). In another study using the same data, researchers reported that Black but not White adults showed a positive association between vigilance and hypertension prevalence. In fact, for Black and White adults who report no vigilant thoughts and behaviors, the inequalities in hypertension prevalence are substantially lower than generally reported in the literature (Hicken et al. 2014) and these remaining inequalities are explained by conventional hypertension risk factors such as smoking, obesity, and heavy alcohol use. However, as vigilance levels increase, the inequalities in hypertension increase substantially and are not explained by any of the conventional hypertension risk factors (Hicken et al 2014). These results suggest that, first, vigilance is particularly salient for Black Americans and, second, chronic vigilance is related to poor health.

We examine the relation between vigilance and three global measures of mental and physical health using this population-representative sample of Black adults in Chicago. We argue that vigilant behaviors represent the same behaviors often discussed within the context of respectability politics. Therefore, this analysis examines an additional potential negative consequence of engaging in respectability politics behaviors.

Methodology

Data

We used data from the Chicago Community Adult Health Study (CCAHS), a cross-sectional survey designed to examine the biological, social, and environmental correlates of adult physical and mental health. The CCAHS is a multi-stage probability sample of 3,105 adults, aged 18 years and older, living in Chicago, stratified into 343 neighborhood clusters. Each neighborhood cluster usually included two census tracts with meaningful physical and social boundaries. Face-to-face interviews were conducted from one respondent per household between May 2001 and March 2003 with a response rate of 71.8%.⁵ While these data were collected 15 years ago, these are the only data available on vigilance and health for a large, population-representative sample of Black adults. Furthermore, while the manifestations of racism have changed over time to fit with broader social mores, racism itself has not abated (Bonilla-Silva 2010; Gainous 2012; Haney-López 2014). Analysis was limited to 1240 non-Hispanic Black respondents (824 women and 416 men).

Variables

Health outcomes

We examined three measures of health that capture different aspects of mental and physical health. Depressive symptoms are thought to capture the psychological distress, misery, or demoralization resulting from social sources (Mirowsky and Ross 2003). Furthermore, researchers argue that measures allowing for gradations in depressive symptoms (rather than those that are indicators of disease) may better capture the mental health variation (resulting from those social sources) in the population (Mirowsky and Ross 2003). Depressive symptoms were measured using an abbreviated version of the Center for Epidemiology Studies-Depression scale (CES-D) (Radloff 1977). Participants were asked how often in the past week they: (1) felt depressed; (2) felt everything was an effort; (3) had restless sleep; (4) was happy; (5) felt lonely; (6) felt people were unfriendly; (7) enjoyed life; (8) did not feel like eating; (9) felt sad; (10) felt that people disliked him/her; and (11) could not get going. Responses were on a 4-pt Likert-like scale indicating: 1=never, 2=hardly ever, 3=some of the time, and 4=most of the time. Responses to the “happy” and “enjoyed life” questions were reverse-coded and a scale was created as the mean of all responses with higher values representing higher levels of depressive symptoms (Cronbach’s alpha=0.85). Self-rated health (SRH), a general overall appraisal of one’s health, has been shown to be related to numerous morbidities and mortality (Idler and Benyamini 1997). While generally captured by a single question, it is thought that the robust relation between SRH and objective measures of health is due to either it: (a) capturing aspects of ill health not detectable biomedically or (b) reflecting broader sociodemographic or psychological determinants of morbidity and mortality (Eriksson, Unden, and Elofsson 2001). SRH was measured with the question, “All in all, would you say that your health is generally excellent, very good, good, fair, or poor?” The responses were coded such that 1=excellent, 2=very good, 3=good, 4=fair, and 5=poor. A count of lifetime chronic conditions captures overall health from multiple systems (e.g., respiratory, cardiovascular) and accounts for co-morbidities. Respondents were asked if a doctor or health care provider had ever told them that they had: (1) a heart attack; (2) other heart trouble; (3) high blood pressure or hypertension; (4) a stroke; (5) chronic bronchitis; (6) asthma; (7) arthritis; (8) diabetes; (9) a stomach, duodenal, or peptic ulcer; (10) poor circulation to the legs or peripheral artery disease; (11) emphysema or chronic obstructive pulmonary disorder; (12) osteoporosis or bones that break easily; or (13) cancer.

Vigilance

Vigilance was assessed using a modified version of the 6-item vigilance scale developed for the 1995 Detroit Area Study (Clark, Benkert, and Flack 2006; Himmelstein et al. 2015). The original scale was based on qualitative research describing how participants anticipated and prepared for racial discrimination (Essed 1990, 1991; Feagin 1991; Feagin and Sikes 1994). We used an abbreviated version of this scale based on responses to the following four questions: In your day-to-day life, how often do you do the following things: (1) try to prepare for possible insults from other people before leaving home; (2) feel that you always have to be very careful about your appearance to get good service or avoid being harassed; (3) try to avoid certain social situations and places; and (4) carefully watch what you say and how you say it. Responses were on a 5-pt Likert-like scale indicating: 0=never, 1=less than once a year, 2=a few times a year, 3=a few times a month, and 4=at least once a week or almost every day. Note that while there was no specific mention of race in the questions, they were asked immediately following the questions about unfair treatment, which contains some reference to race, as well as other social factors such as gender, appearance, and age. When responses a summed to create a continuous scale, the Cronbach’s alpha=0.74. However, we operationalized the scale to reflect important level differences among those who report never, sometimes, and frequently engaging in vigilant thoughts and behaviors as follows: those who reported “never” on all four items were coded as never=0; those who reported “at least once a week or almost everyday” on two of the items or “a few times a month” on all four items were coded as frequently=2; and all others were coded as sometimes=1. This operationalization of vigilance reflects the notion that regular and frequent vigilance is important for health and that never engaging in any vigilant thoughts and behaviors has important implications for health as well.

Covariates

Respondents were asked their race and ethnicity based on U.S. Office of Management and Budget (OMB)-designated racial (White, Black, American Indian, Asian, Pacific Islander) and ethnic (Hispanic, not Hispanic) categories. Race was categorized as non-Hispanic White, non-Hispanic Black, Hispanic, and non-Hispanic other (which included American Indian, Asian, and Pacific Islander). We include only those who were categorized as non-Hispanic Black. We included two measures of SES to control for potential confounding. Education was measured from participant reports of the number of years of education completed. Household poverty income ratio (PIR) was calculated as the ratio of household income to the 2001–2003 average USDA poverty threshold for the household size. Higher PIR values indicate greater relative household income. Additional covariates include gender (female = 1) and age (range: 18 to 92) in years. Education, household poverty, and gender do not likely confound the relation between vigilance and health in this sample because, while they are all related to health, they are not related to vigilance in unadjusted models. However, we include them in addition to age (which is a confounder) as they are related to health and will provide a conservative estimate of the relation between vigilance and health.

Analytic Approach

To examine the sociodemographic, vigilance, and health descriptive characteristics, we estimated means, with standard errors, for continuous variables, and percentages for categorical variables for the total sample, by gender, and by level of vigilance, using t-tests to test for differences by gender and level of vigilance. We include standard errors rather than standard deviations because our data are multiply imputed, as described below.

To address our research question on the association between vigilance and health, we estimated models using linear (for depressive symptoms), ordinal logistic (for SRH), and negative binomial (for chronic conditions) regression, adjusting for age, gender, education, and household PIR. We employed negative binomial regression because the variance of the chronic conditions distribution is nearly double the mean (mean=1.19; variance=2.35) in this sample, indicating slight overdispersion. Furthermore, because chronic conditions are reported for one’s lifetime and the sample is composed of adults of a wide age range, we adjusted for the exposure time to develop these conditions as age in years.

We performed several sensitivity analyses. First, because the chronic conditions scale requires a doctor or health care provider diagnosis and regular health care access may vary by vigilance (Abdou et al. 2016; Aronson et al. 2013; Burgess et al. 2010) we adjusted models with a four-item health care access scale. This scale includes questions about whether the respondent has visited a health care provider and had blood pressure or cholesterol checked within the previous two years. While this measure is positively related to the chronic conditions outcome, its inclusion did not change the relation between vigilance and chronic conditions at all (results available upon request). Second, because the chronic conditions scale includes items which are strongly age-related and because vigilance may result in earlier onset of these conditions, we modeled the interaction between vigilance and age. The interaction between vigilance was zero, indicating that the relation between age and chronic conditions does not vary by level of vigilance (results available upon request).

There was missing information on PIR (due to missing information on household income, n=242) and vigilance (n=3); these data were multiply imputed using IVEware (University of Michigan, Ann Arbor, MI) via SAS (SAS Institute, Cary, NC) to create five imputed datasets. We used the multiple imputation suite of commands in STATA, which “adjusts coefficients and standard errors for the variability between imputations according to the combination rules by Rubin” (Rubin 1987; StataCorp 2015, 41) to analyze the imputed data. All analyses were weighted to account for complex survey design, differential selection into the sample, non-response, and household size. With respect to race/ethnicity, age, and gender, the distribution of the weighted sample and the 2000 Census estimates were comparable. All analyses, with the exception of the creation of the multiply-imputed dataset, were conducted in STATA 14.1MP (StataCorp; College Station, TX).

Results

While the vigilance scale has a Cronbach’s alpha=0.74, meaning that the four items of the scale are related to each other in our Chicago sample, the response patterns of the individual items vary (see Table 1). In response to the two questions asking about preparation for insults and taking care of one’s appearance, most adults report that they have never engaged in these behaviors (“insults”=69%; “appearance”=60%; see Table 1). Nevertheless, roughly 14–15% of adults report engaging in these behaviors at least once a week (“insults”=14%; “appearance”=15%; see Table 1). On the other hand, in response to the two questions asking about watching what one says and avoiding social situations, most adults report either never engaging in these behaviors or engaging in them at least once a week. Specifically, 28% and 34% of adults report never watching what they say or avoiding social situations, respectively, while 44% and 29% report these respective behaviors at least once a week (see Table 1). These patterns are consistent across gender, however, more men than women report taking care of one’s appearance (p=0.005; see Table 1). When examining all vigilance items together, 18% of adults report never engaging in any of these behaviors, while 31% are engaging in some aspect of vigilance frequently. This pattern is consistent across gender.

Table 1.

Distribution of responses to vigilance survey questions by gender, Chicago Community Adult Health Study 2000–2002

	Total	Women	Men	pWvM
In your day-to-day life, how often do you do the following things:
Prepare for possible insults before leaving home				0.492
Never	69	70	67
<Once/year	9	7	11
Few times/year	5	5	5
Few times/month	4	4	4
Almost everyday	14	14	14
Be very careful about your appearance				0.005
Never	60	65	54
<Once/year	8	8	9
Few times/year	9	7	12
Few times/month	7	7	7
Almost everyday	15	13	18
Carefully watch what you say and how you say it				0.305
Never	28	28	28
<Once/year	7	7	8
Few times/year	13	13	14
Few times/month	8	7	9
Almost everyday	44	46	41
Try to avoid certain social situations and places				0.530
Never	34	36	33
<Once/year	10	10	9
Few times/year	15	13	18
Few times/month	12	14	9
Almost everyday	29	28	31
Composite vigilance scale				0.496
Never	18	18	18
Sometimes	51	52	49
Frequently	31	30	33

Notes: Values shown are percentages. p_WvM is the p-value of the t-test for the difference in responses between women and men. The entire question wording is: “Try to prepare for possible insults from other people before leaving home”; “Feel that you always have to be very careful about your appearance to get good service or avoid being harassed”; “Carefully watch what you say and how you say it”; “Try to avoid certain social situations and places”.

The variation in the frequency of engaging in specific vigilant behaviors suggests that presentation of self strategies used by Black Americans in this sample are more likely to be based on oral communication versus style of dress. In addition, strategies to mitigate the experience of discrimination may be based more on avoidance versus preparation for bias. However, it is important to note that a significant proportion of individuals engaged in at least one of the four vigilant behaviors and that these behaviors are related.

Of the nine chronic conditions included in the scale, 35% of adults report none of these conditions while 22% report three or more of these conditions and this varies across gender (see Table 2). Seventeen percent of men but 25% of women report three or more chronic conditions (p=0.010 for gender differences; see Table 2). Depressive symptoms can range from one (never) to four (most of the time) and have a mean of 1.92 (se=0.02; see Table 2), which is consistent across gender (p=0.085 for gender difference; see Table 2). That there are gender differences in high numbers of chronic conditions but not in depressive symptoms reflects in the lack of gender differences in SRH, which is thought to have a strong mental health component (Idler and Benyamini 1997). In the overall sample, 48% of adults report very good or excellent health and three percent report poor health (see Table 2; p=0.085 for gender difference). Notably as well, while a large proportion of the men and women in this sample report three or more chronic conditions, only a small number report poor health, suggesting perhaps that these chronic conditions are not acutely life-threatening and simply accepted as part of life or are seen as being more manageable and, therefore, may alter the parameters of how individuals in the sample respond to the SRH question (Robertson et al. 2009).

Table 2.

Health and demographic characteristics by gender, Chicago Community Adult Health Study 2000–2002

	Total	Women	Men	pWvM
Chronic health conditions, count (%)				0.010
0	35	32	39
1	27	28	26
2	15	14	17
3	11	12	9
4	5	6	3
5	4	4	4
6	1	1	1
7	1	1	0
8	0	1	0
9	0	0	0
Depressive symptoms				0.085
	1.92 (0.02)	1.95 (0.03)	1.88 (0.03)
Self-rated health, (%)				0.082
Excellent	14	13	15
Very good	34	32	36
Good	34	34	34
Fair	15	17	13
Poor	3	4	3
Female, (%)	58
Age, years	44.17 (0.58)	44.49 (0.68)	43.73 (0.99)	0.527
Education, years	12.79 (0.11)	12.89 (0.14)	12.66 (0.16)	0.244
Household PIR	2.58 (0.20)	2.21 (0.11)	3.07 (0.43)	0.046

Notes: Values represent means with standard errors in parentheses, unless otherwise noted. p_WvM is the p-value of the t-test for the difference between women and men. Percentages may not sum to those in Table 3 due to rounding. Abbreviations: PIR, poverty income ratio.

Without adjusting models for sociodemographic characteristics, vigilance is not related to chronic conditions, but is related to depressive symptoms and SRH (see Table 3). After adjustment for age, gender, education, and household PIR, vigilance is related to all three health outcomes. This suggests that, for chronic conditions, but perhaps less so for the other health outcomes, these sociodemographic characteristics may be differentially related to vigilance and chronic conditions such that, without adjusting for them, any relations cancel out. For example, younger adults report greater levels of vigilance but also have lower levels of chronic conditions compared to older adults (results available upon request). Thus the vigilance-health relation may be masked without adjustment for age.

Table 3.

Health and demographic characteristics by level of vigilance, Chicago Community Adult Health Study 2000–2002

	Never	Some	Freq	p0v1	p0v2
Chronic health conditions, count (%)				0.807	0.897
0	35	35	36
1	25	28	25
2	14	16	15
3	13	9	13
4	4	5	3
5	5	4	3
6	2	1	2
7	1	0	2
8	0	1	0
9	0	0	0
Depressive symptoms				0.000	0.000
	1.66 (0.04)	1.93 (0.03)	2.06 (0.04)
Self-rated health, (%)				0.055	0.003
Excellent	17	13	11
Very good	37	34	31
Good	31	34	34
Fair	12	15	19
Poor	2	3	4
Female, (%)	58	59	55	0.863	0.592
Age, years	48.35 (1.50)	43.76 (0.85)	42.41 (1.06)	0.010	0.002
Education, years	12.54 (0.23)	13.04 (0.15)	12.53 (0.21)	0.061	0.977
Household PIR	2.58 (0.22)	2.83 (0.36)	2.16 (0.16)	0.541	0.107

Notes: Vigilance response categories are: Never, Sometimes, and Frequently. Values represent means with standard errors in parentheses, unless otherwise noted. p_0v1 and p_0v2 are the p-values of the t-tests for the difference between the never and sometimes and frequently categories, respectively. Percentages may not sum to those in Table 2 due to rounding. Abbreviations: PIR, poverty income ratio.

Compared to those who report never engaging in vigilant behaviors, those who frequently engage in these behaviors show a 25% greater lifetime count of chronic conditions (incident rate ratio, IRR=1.25; 95% confidence interval, CI: 1.05, 1.48; see Table 4). For example, if those who report never engaging in vigilance behaviors ultimately report four chronic conditions over the course of a lifetime (e.g., asthma in childhood and then diabetes, hypertension, and arthritis in adulthood), those who report high levels of vigilance would report one additional condition in their lifetime.

Table 4.

Association between vigilance and health, Chicago Community Adult Health Study 2000–2002

	Chronic Conditions IRR(95%CI)	Depressive Symptoms b(95%CI)	Self-Rated Health OR(95%CI)
Vigilance
Never	Ref	Ref	Ref
Sometimes	1.17† (0.99, 1.40)	0.27*** (0.18, 0.35)	1.67** (1.18, 2.31)
Frequently	1.25** (1.05,1.48)	0.38*** (0.29, 0.48)	1.98*** (1.38, 2.82)
Age	1.01*** (1.01, 1.03)	0.00** (−0.01, 0.00)	1.02*** (1.01, 1.03)
Gender	1.20* (1.03, 1.41)	0.08† (−0.01, 0.17)	1.28† (0.97, 1.69)
Education	0.97** (0.95, 0.99)	−0.03** (−0.04, −0.01)	0.88*** (0.84, 0.93)
Household PIR	0.96* (0.92, 0.99)	−0.01 (−0.02, 0.01)	0.98 (0.92, 1.02)
Intercepts and cuts
Intercept or cut 1		2.02*** (1.74, 2.30)	−2.16*** (−3.05, −1.27)
Cut 2			−0.30 (−1.16, 0.56)
Cut 3			1.38** (0.52, 2.21)
Cut 4			3.38*** (2.54, 4.22)

Notes:

^***p<0.001;

^**p<0.01;

^*p<0.05;

^†p<0.10. Notations of significance may not match the 95% confidence intervals due to rounding of interval values. For example, the association between education and chronic conditions is significant at the p<0.05 level even though the upper value of the 95% confidence interval is 1.00. The actual value to three decimal places is 0.996. “Cuts” refers to the results of ordinal logistic regression for self-rated health and is the estimated cutpoint between levels of self-rated health when all covariates are held at their base levels.

Vigilance is related to depressive symptoms in a graded manner. Compared to those who report never engaging in vigilant behaviors, those who sometimes engage in these behaviors report a greater level of depressive symptoms; those who frequently engage in these behaviors report an even greater level of depressive symptoms (see Table 4). After adjusting for sociodemographic characteristics and vigilance, the mean level of depressive symptoms is about a two on a scale of zero to four (Intercept=2.02; 95%CI: 1.74, 2.30; see Table 4). Compared to those who report no vigilance, those who report regular vigilance show a depressive symptoms level that approaches half a point greater (b=0.38; 95%CI: 0.29, 0.48; see Table 4). While this may not seem like much at the individual level, small shifts like this make a large difference at the population level (Rose 2001). Furthermore, consider this in comparison with education, a well-established determinant of mental health with those with higher education reporting lower depressive symptoms than those with lower education (Lorant et al. 2003). Compared to those adults who have a post-baccalaureate education (e.g., graduate or professional school), those who have not reached the ninth grade report, a substantial difference in educational level, show a quarter of a point greater depressive symptoms level (b=0.25; 95%CI: 0.04, 0.48; results not shown in table form).

Finally, vigilance is also inversely related to SRH such that more frequent engagement in vigilant behaviors is related to poorer SRH (see Table 4 for odds ratios). For clarity of interpretation, we estimated the predicted probabilities of each level of SRH for each level of vigilance (see Table 5). For example, 21% of those who report never engaging in vigilant behaviors also report excellent SRH, while only 12% of those who frequently engage in these behaviors report in excellent SRH. On the other hand, 13% of those who report never engaging in vigilant behaviors report fair or poor SRH while 22% of those who frequently engage in these behaviors report this level of SRH.

Table 5.

Predicted probability of self-rated health categories by level of vigilance, Chicago Community Adult Health Study 2000–2002

	Never	Some	Freq
Excellent	21	14	12
Very good	38	34	32
Good	28	33	34
Fair	11	16	18
Poor	2	3	4

Notes: Values are predicted probabilities in percentage. Vigilance response categories are: Never, Sometimes, and Frequently.

Discussion/Conclusion

In summary, we find that engaging in vigilant behaviors is common for a majority of Black Americans across genders. This finding highlights the somewhat ubiquitous nature of these behaviors among Black Americans as a necessary strategy to navigate everyday life in a racially hierarchical society. However, there is still variation in these behaviors, with some Black Americans engaging in these behaviors almost daily while others engaging in these behaviors only some of the time and a small percentage never engaging in these behaviors. While it is beyond the scope of this analysis to determine factors that are associated with intensity of engagement in these behaviors or choice of specific behaviors, it is clear, that, like debates around respectability politics some Black Americans employ these behaviors more than others, but a majority of African Americans at least employ these behaviors some of the time.

Consistent with prior research, we find that a substantial proportion of Black Americans report three or more chronic health conditions, reflecting the disproportionate burden of chronic diseases among African Americans compared to whites (Centers for Disease Control and Prevention 2011, 2013). Moreover, these are likely underestimates of disease burden as Black Americans are more likely to be unaware of the chronic conditions that they have (Chatterji, Joo and Lahiri 2012). Reports of fair or poor health status and depressive symptoms also mirror national statistics and capture health burdens faced by this population (Centers for Disease Control and Prevention 2011, 2013; George and Lynch 2003).

The most important contribution of our study is in providing evidence based on a representative sample that engaging in vigilant behaviors and thoughts, especially frequently, is associated with increased poor physical and mental health as indexed by chronic health conditions, depressive symptoms, and self-rated health. This study builds on previous research evidence, which points to the powerful connections between vigilance and health status measured by hypertension and sleep problems (Hicken et al. 2013; Hicken et al. 2014). Given that vigilant behaviors are endorsed by proponents of respectability politics (but likely engaged in by a much broader swath of Black Americans), our findings suggest that respectability politics discussions need to consider the health consequences of negotiating racialized everyday social spaces as a Black American. Indeed, our findings also suggest the ways in which engagement in these thoughts and behaviors may be a major determinant of the persistent racial/ethnic health inequalities that exist in the United States. In other words, given that a majority of Black Americans are engaging in these behaviors at least some of the time, and that engaging in these behaviors, especially frequently, poses a significant health risk, then it is possible that these behaviors contribute to large differences we see in health between Blacks and Whites, who are not engaging in these behaviors with the same frequency, if at all. Restated, vigilance may serve as a unique racialized stressor serving to disproportionately impact the health of African Americans. It is beyond the scope of this analysis to formally test this hypothesis but this is an important direction for future research.

Although we are the first to examine vigilant behaviors as a proxy for Black respectability politics behaviors, previous work has also identified potential negative consequences to believing in meritocratic values in the face of racial discrimination (Hagiwara, Alderson, and McCauley 2015; Kwate and Meyer 2010). Some have argued that this, too, may be an ideology held by those who endorse Black respectability politics or, at least, beliefs that may inform Black respectability politics. Moreover, that these beliefs can also make you sick (Kwate and Meyer 2010; Womack 2016). We see our work as expanding this area of research by directly testing the behaviors that align with these beliefs.

Future work should further unpack gender differences in the health consequences of vigilance. In this study we focused on three measures of global health. However, research suggests that the relation between chronic stress and health are gendered (Banks and Kohn-Wood 2002; Perry, Harp, and Oser 2013; Umberson et al. 2014). For example, we may find that vigilance may be more consequential for influencing body mass index for women than men, as women are more likely to overeat and engage in sedentary behaviors to cope with stress than men (Jackson, Knight, and Rafferty 2010; Mezuk et al. 2013). Future research should also employ the use of longitudinal data, preferably using nationally representative data, to examine the associations between vigilance and health to better clarify the temporal ordering that would underlie causal associations. Our current analysis can only provide evidence of an association between vigilance and health for adults living in Chicago. Furthermore, we examine cross-sectional associations between vigilance and health and there is a possibility that either health status affects the report of vigilant behaviors or that there is an important omitted confounder that links reports of these behaviors to health. Future work should examine these associations as well using longitudinal data.

Future work should also consider exploring other forms of vigilant behaviors that may also be salient to health and also related to respectability politics. For example, the measure of avoidance behavior used in this analysis likely underestimates the ways in which African Americans must navigate social spaces in order to avoid discrimination. Because the possibility of discrimination may be omnipresent in all social spaces, African Americans may need to engage in other behaviors in order to avoid harassment. For example, Anderson’s (1990) qualitative study conducted from 1975 to 1989 in an urban community, describes the ways in which Black Americans must actively prepare for discriminatory experiences and avoid their escalation. He found that young Black men in his sample felt that they must carry some form of identification to prepare for possible encounters with police, particularly in certain social spaces, to avoid arrest or unwarranted suspicion. One young African American male former college student noted the need to be prepared for police encounters when travelling through campus:

I know I [used] to feel when I was enrolled in college last year, when I had an ID card. I used to hear stories about the blacks getting stopped over by the dental school, people having trouble sometimes. I would see that all the time. Young black male being stopped by the police. Young black male in handcuffs. But I knew that because I had that ID card I would not be mistaken for just somebody snatching a pocketbook, or just somebody being where maybe I wasn’t expected to be (Anderson 1990, 198).

Recent media attention on multiple incidents of police violence against unarmed black men, women and, even children, also highlight the extra steps African Americans take to in order to try to somehow protect against or avoid unwarranted and unfair treatment (from the minor to the extreme) in everyday social spaces that must be navigated in order to live, learn, work and play. At the same time these incidents call into question the usefulness of these strategies for avoiding discrimination, or in these situations, getting killed (Gandbhir and Foster 2015; Zavadski 2014). Moreover, our results suggest that engagement in these behaviors is also associated with additional mental and physical health costs.

In one final and related note, we offer that vigilant thoughts and behaviors are natural anticipatory coping responses to life in a racist environment. Black Americans should not be blamed or condemned for engaging in these behaviors. We do not label these behaviors as “right” or “wrong,” as has been done in the popular literature. However, we propose that rather than expending time and energy debating and moralizing these behaviors, that we ask what in our American social structure and culture demands that Black Americans engage in these coping vigilant behaviors (Ryan 1976). In other words, we propose to maintain focus on the aspects of American society that maintain our racially-hierarchical and racialized structure and culture that burdens Black Americans with the need to chronically engage in vigilant behaviors.