Fear of a Black Neighborhood: Anti-Black Racism and the Health of White Americans

Abstract

Does anti-Black racism harm White Americans? We advance hypotheses that address this question within the neighborhood context. Hypotheses are tested with neighborhood and survey data from a probability sample of White residents of Nashville, Tennessee. We find that regardless of neighborhood crime rates or socioeconomic compositions, Whites report heightened perceptions of crime and danger in their neighborhoods as the proportion of Black residents increases. Perceived neighborhood danger, in turn, predicts increased symptoms of psychophysiological distress. When stratified by socioeconomic status (SES), however, low-SES Whites also report perceptions of higher status when living near more Black neighbors, which entirely offsets their distress. We conclude that although anti-Black racism can ironically harm the health of White Americans, compensatory racist ideologies can also offset these harms, particularly for lower-status Whites. We situate our findings within broader discussions of anti-Black racism, residential segregation, and psychiatric disorders commonly observed among White Americans.

Introduction

Although decades of research have established the toxic effects of racism on people racialized as Black (hereafter Blacks), little is known about whether or how racism harms people racialized as White (hereafter Whites). Despite clear advantages, Whites have long exhibited higher rates of psychiatric disorder and suicide relative to Blacks—rates that appear to be accelerating, especially for lower-status Whites (Barnes and Bates 2017; Case and Deaton 2015, 2017; Siddiqi et al. 2019). Whites also appear vulnerable to distress in racially integrated spaces (Mendes et al. 2007; Sampson and Raudenbush 2004; Steele 2021; Yancey 2008). Patterns like these have led some scholars to infer that racism reflects a damaged social order (Lucas 2008), one that can even harm Whites in some contexts (Malat, Mayorga-Gallo, and Williams 2018; McGhee 2021; Metzl 2019).

This study asks the question: How does anti-Black racism harm Whites? We address this question within the neighborhood context, focusing on how fear of Blackness triggers chronic stress for Whites in racially diverse communities. Merging neighborhood and survey data from a Nashville-based sample, we find that an increasing presence of Black neighbors predicts heightened perceptions of neighborhood danger and related distress among White residents, regardless of neighborhood crime rates or socioeconomic compositions. However, lower-status Whites also derive a sense of higher status from living near more Black neighbors, which entirely buffers their distress.

Our study offers critical new insights into racism-related disparities in the United States. We advance racism research by showing how anti-Black racism can ironically harm Whites by triggering unwarranted fears in racially diverse spaces. This finding, in turn, builds on a longstanding body of research that highlights residential segregation as a social determinant of racialized inequities (Du Bois 1899; Williams 2012). Departing from this literature, our study contributes to newer work that has revealed strained relations between residents of racially integrated spaces (DeAngelis, 2022; DeAngelis, Hargrove, and Hummer 2022; Mayorga-Gallo 2014). Finally, our study may also have broader implications for declining mental health among Whites (Case and Deaton 2015, 2017), trends that some scholars are attributing to a growing sense of racial status threat in an increasingly diverse and globalizing society (Rambotti 2022; Siddiqi et al. 2019).

In what follows, we first provide a historical overview of fear of Blackness among Whites. We then develop a series of hypotheses for how fear of Blackness can affect Whites’ health in certain neighborhood and socioeconomic contexts. After this, we present our conceptual model, data, methods, and findings. We close by considering some broader implications and limitations of our study, as well as avenues for future research.

Background

Fear of blackness

Following prior work in this area, we define fear of Blackness as the phobic reaction to dark-skinned persons or groups (Fanon 1952; Jones and Obourn 2014; Steele 2021). In this first section, we consider how fears of Blackness have served the collective interests of Whites in upholding White supremacy and stark racialized boundaries. This discussion will provide historical context for understanding why Whites still tend to associate Blackness with fearful imageries.

For over a century, Black scholars have vividly portrayed how fears of Blackness constitute the very foundations of White identity and history (Alexander and Alexander 2021; Du Bois 1899; Fanon 1952; Yancey 2008). In Black Bodies, White Gazes, George Yancy (2008) discusses how Black bodies have functioned historically as ideological construction sites for an illusory White self-concept, whose sense of freedom could be conceived only in relation to dehumanized Black subjects (see also Mills 1997). In their aptly titled contribution to the 1619 Project, “Fear,” Alexander and Alexander (2021) demonstrate that much of US history can be framed around Whites’ fearful reactions to Blacks’ struggles for freedom and equity.

Fears of Blackness have deep historical roots. Looming threats of enslavement first motivated colonial-era European labor coalitions to construct a White identity in contradistinction with enslaved Africans, whose dark skin could be exploited as a palpable demarcation between “free White labor” and “Black chattel” (Morgan 1975; Roediger 1991). Later waves of European immigrants would then uphold anti-Blackness by fighting to “become White” in efforts to secure the privileges of US citizenry for themselves and their progeny (Ignatiev 2012; Lopez 2006). To this day, many Whites still appear to define themselves and their communities in similar relational manners, going so far as to embrace policies that undermine their own well-being if it means dissociating from the perceived inferiorities of Blackness (Metzl 2019).

Fears of Blackness are most visible today in Whites’ attitudes toward crime and punishment, and within the broader criminal justice system. Although Whites are much less likely than Blacks to be victims of crimes (Dixon and Linz 2000), Whites still express greater fears of victimization and support harsher punishments for persons convicted of crimes, most of whom Whites assume to be Black (Ghandnoosh 2014). It is perhaps unsurprising, then, that Blacks are grossly overrepresented among incarcerated populations (Alexander 2010), as well as within databases of fatal police encounters (Edwards, Esposito, and Lee 2018), despite evidence suggesting Blacks are less likely than Whites to be armed or show signs of distress during such encounters (DeAngelis 2021).

Legacies of institutionalized segregation further betray pervasive fears of Blackness among Whites. Through interrelated practices of racial covenants, blockbusting, and redlining, real estate markets have long profiteered from Whites’ fears that Black neighbors signaled inevitable property decline, social disharmony, and crime (Rothstein 2017). Whites express similar misgivings in the post-Civil Rights era, albeit through subtler or “color-blind” mannerisms (Bonilla-Silva 2003). For example, surveys find that Whites still prefer to live in majority or all-White communities, associating Black neighbors with threats to neighborhood safety and status (Bobo and Zubrinsky 1996; Lewis, Cogburn, and Williams 2015; Emerson, Chai, and Yancey 2001). When asked to justify their preferences, however, Whites rarely reference race, pointing instead to better schools and amenities in White communities (Johnson and Shapiro 2003; Roda and Wells 2013).

In short, Whites have promulgated fearful imageries of Blackness for centuries, employing various sociopolitical mechanisms to distance themselves from Blacks (Bonilla-Silva 2019; Feagin 2020). Thus, although fears of Blackness are clearly psychic in nature, these emotions have also served the collective interests of Whites in upholding Black–White boundaries (Roediger 1991; Rothstein 2017; Winling and Michney 2021). Given this historical context, we now consider how Whites might react whenever these boundaries are breached.

Vigilance hypothesis

George Yancy describes an encounter he once had inside an elevator with a White woman, who appeared visibly disturbed by his presence as a Black man. He imagines what she must have felt as they stood beside each other:

Her palms become clammy. She feels herself on the precipice of taking flight, the desperation to flee. There is panic, there is difficulty swallowing, and there is a slight trembling of her white torso, dry mouth, nausea (Yancey 2008, 21).

Yancey is describing what is often labeled an acute “fight-or-flight” response. In support of Yancey’s observation, experimental evidence indicates that some Whites experience increased amygdala activity, a subcortical structure that plays a role in emotional learning and conditioned fear, when viewing unfamiliar Black faces (Phelps et al. 2000). Other studies have documented similar physiological responses, such as increased heart rate, when Whites encounter Black strangers (Vrana and Rollock 1998). Evidence suggests these stress responses are likely triggered by Whites’ implicit associations of Blacks with violence and criminality (Eberhardt et al. 2004).

We conceptualize distress from fears of Blackness as an inverted form of racism-related vigilance for Whites, or what we term “anti-Black vigilance.” To date, research on racism-related vigilance has exclusively focused on how anticipatory stress over future racist encounters harms the health of Blacks, especially for those who regularly navigate predominantly White spaces (DeAngelis 2022; Hicken et al. 2013, 2014; Sewell et al. 2016; Smith, Allen, and Danley 2007). This body of work shows how constant fears of anti-Black discrimination can tax psychophysiological systems and lead to exhaustion and distress.

From the converse perspective, we conceptualize anti-Black vigilance among Whites as the propensity to monitor social spaces for racialized “intruders.” Anderson (2015, 13) has offered a general depiction of anti-Black vigilance, noting that when an unknown Black person enters a social space with White people, Whites will often “immediately try to make sense of him or her—to figure out ‘who that is,’ or to gain a sense of the nature of the person’s business and whether they need to be concerned.” Feagin and Sikes (1994) also interviewed over 200 upper-middle-class Blacks, who reported that White neighbors regularly monitored their behaviors and movements in tandem with local police. Other studies have revealed that Blacks tend to report more discrimination as the proportion of White residents in their neighborhoods increases (DeAngelis 2022).

Given the historical links between Whites’ preferences for residential isolation and fears of Blackness, we examine how anti-Black vigilance generates stress for Whites within racially diverse neighborhoods. Links between the neighborhood context and resident health are well-established (Hill and Maimon 2013). Moreover, research finds that neighborhood racial context is a unique predictor of resident stress and health (DeAngelis 2022; Hutchinson et al. 2009). Although most research in this area documents health correlates of racial composition for Black residents (English et al. 2014; Hurd et al. 2013; Hutchinson et al. 2009; Reichman, Teitler, and Hamilton 2009; Vogt Yuan 2007), some research suggests racial composition also matters for Whites’ health (Fang et al. 1998; Halpern and Nazroo 2000; Henderson 2005; Inagami et al. 2006; LeClere, Rogers, and Peters 1997; Li, Wen, and Henry 2014). Intriguingly, researchers have observed patterns of increased mortality (Fang et al. 1998; Inagami et al. 2006; LeClere, Rogers, and Peters 1997), obesity (Li, Wen, and Henry 2014), and mental health problems among White residents in predominantly Black areas (Henderson 2005).

Noticeably absent from this literature is an examination of psychosocial mechanisms linking neighborhood racial composition to the health of White residents. Toward this end, we find parallel insights from criminology to be instructive. Research in this area finds that Whites who live in neighborhoods with a higher proportion of Black residents tend to report increased perceptions of crime and fears of victimization, oftentimes regardless of neighborhood crime rates (Chiricos, Hogan, and Gertz 1997; Quillian and Pager 2001). Moreover, studies find that such perceptions predict greater psychological distress (Ross 1993), depression (Stafford, Chandola, and Marmot 2007), and anxiety (Grinshteyn et al. 2017).

Studies reviewed in this section, as well as others in the broader stress process literature (Cohen and Janicki-Deverts 2012; Geronimus et al. 2010; Pearlin 1999), support our contention that fears of Blackness and anti-Black vigilance could serve as chronic stressors for Whites in racially diverse neighborhoods. Moving forward, we will refer to this expectation as the “vigilance hypothesis.” In the following sections, we consider how the mental health consequences of living near more Black neighbors may vary for Whites depending on their socioeconomic status (SES). We then summarize our hypotheses and introduce our data and findings.

Comparison hypothesis

The health consequences of living near more Black neighbors could vary for Whites based on their SES. Although all Whites have benefited from anti-Black racism, lower-SES Whites appear to have been the most vocal guardians of White supremacy throughout US history (Alexander and Alexander 2021). Du Bois was the first to suggest that low-status Whites derive a compensatory sense of prestige from vicariously overseeing the oppression of Black persons—what he termed “wages of Whiteness.” In Black Reconstruction, for instance, Du Bois observed that poor Whites benefited psychosocially from serving on local slave patrols, as their service “fed their vanity” and “associated them with the masters” (Du Bois 1935, 81). Roediger (1991) would later elaborate on this process, revealing how White laborers have built a collective sense of esteem by projecting their insecurities onto oppressed Black groups throughout US history.

The “wages of Whiteness” hypothesis entails a complex relational process comprising subjective status appraisals vis-à-vis racial and SES reference groups. We draw from the social comparison principle in psychology to derive hypotheses for potential status appraisal processes (Festinger 1954). This principle states that people derive social esteem by comparing their current statuses and achievements with salient reference groups. Moreover, people typically make two kinds of comparisons based on their chosen referent: downward or upward. Downward comparisons involve contrasting oneself with a perceived lower-status group, whereas upward comparisons involve contrasts with a higher-status group (Lockwood 2002; Wills 1981).

We refer to “wages of Whiteness” as the downward comparison hypothesis. Accordingly, lower-status Whites could generate positive status appraisals—and, ultimately, improved health—by performing downward comparisons with Black neighbors, who represent a marginalized referent. To paraphrase Cheryl Harris, this does not mean that lower-status Whites will perceive themselves as “winners,” but rather “not losers,” assuming that “losing is defined as being on the bottom of the social and economic hierarchy—the position to which Blacks have been consigned” (Harris 1993, 1759). In stress process terms, perceptions of higher racial status could ultimately offset distress for low-SES Whites who live near more Black neighbors.

Contemporary evidence is consistent with the downward comparison hypothesis. In a 2000 book, The Dignity of Working Men, Lamont interviewed Whites of varying occupations, revealing that professional and managerial Whites compared their achievements to “people like themselves, that is, white middle class people who are highly educated,” whereas working-class Whites were inclined to contrast their lots with Blacks (Lamont 2000). Crucially, working-class Whites also appeared to derive feelings of prestige from perceiving themselves as more ambitious and hard-working than their Black peers, who were viewed as lazy and morally inferior (Lamont 2000, 56). In a more recent study, McDermott, Knowles, and Richeson (2019) interviewed working-class Whites from the Midwest. They found that nearly half of participants expressed anti-Black sentiments grounded in the misperception that Blacks always expect handouts from the broader society, contrary to Whites.

The General Social Survey (GSS) provides further evidence of ongoing wages of Whiteness.¹ Since the 1970s, the GSS has asked Whites to assess hypothetical explanations for Black-White disparities in the United States. One explanation is that “most Blacks just don’t have the motivation or will power to pull themselves up out of poverty.” In 2021, 43% of Whites with less than high school education agreed with this statement, compared with just 20% of their college-educated peers. When split by subjective status, 42 and 30% of Whites who identify as “low” or “working” class agreed, respectively, relative to 22% of their “high class” White peers. These broader patterns indicate that lower-status Whites across the country are still more inclined to endorse racist ideologies denigrating Blacks for their supposed lack of work ethic.

Some evidence suggests that higher-status Whites may also derive a status boost from living near more Black neighbors. For example, one study found that higher-status Whites who lived in racially diverse areas felt they were demonstrating progressive values, and thereby expressed feelings of superiority relative to Whites who lived in less diverse areas (Mayorga-Gallo 2014). However, due to longstanding sociopolitical circumstances that have made anti-Blackness more salient to working-class White identity, we argue that the downward comparison hypothesis will be relevant mostly for low-status Whites (e.g., Roediger 1991; Metzl 2019).

There is also some evidence to suggest the contrary of the downward comparison hypothesis. According to the upward comparison hypothesis, lower-status Whites who live near more Black neighbors may compare themselves to high-status Whites, rather than lower-status Blacks, and thus experience even greater distress. One study found that lower-status Whites in the Deep South of the United States, where our study takes place, identified primarily with being White rather than with their SES. This ultimately meant that Whites who resided “in a working-class community situated near black neighborhoods conferred a sense of failure for having not lived up to the affluent, suburban, privileged connotations of whiteness” (McDermott 2006, 2). Another study of Nashville found that Whites who resided in disadvantaged Black neighborhoods reported lower subjective status than their peers in higher-status White areas (DeAngelis 2022). The author of that study also suggested that “residing in [predominantly Black] neighborhoods, isolated from their high-status White compatriots, may be interpreted as a sign of failure” (DeAngelis 2022:1526).

Conceptual model

Our conceptual model is depicted in Figure 1. First, the vigilance hypothesis suggests that an increasing presence of Black neighbors will predict greater distress via perceptions of neighborhood danger. Second, the comparison hypothesis suggests that more Black neighbors could also trigger downward or upward social comparisons, reflected by higher or lower subjective status, respectively. In turn, higher subjective status could counteract or buffer distress from an increased sense of danger, whereas lower subjective status could amplify distress. The comparison hypothesis also entails that individual SES will moderate the stress processes in Figure 1 (not shown). As we explain below, we test for moderation by estimating multigroup models stratified by SES.

Figure 1.

Conceptual model.

Data and Methods

Data

Data come from the Nashville Stress and Health Study (NSAHS), a cross-sectional probability survey of US-born, non-Hispanic Black and White adults who lived in Davidson County, Tennessee between 2011 and 2014 (n = 1252). The NSAHS was collected in two stages. In the first stage, a simple random sample identified 199 block groups, from which 7000 randomly selected addresses were sampled and screened. Of the households screened, 3028 were deemed eligible for the study. From these households, four stratified random samples were drawn of individuals aged 25 to 65, with equal representations of Black and White women and men. Fifty-nine percent of contacted persons ultimately agreed to participate in the study. All analyses are weighted for the probability of selection during the household screening phase, and for nonresponse during the interviewing phase. Poststratification weights are incorporated into the final design weight to permit generalizability to Davidson County’s population of Black and White working-age adults.

All survey interviews were computer-assisted and conducted in the respondent’s home or on the Vanderbilt campus. Interviewers were trained and matched to respondents based on race. Average interviews lasted around 3 hours. Our main analyses include only the subsample of White respondents (n = 625). Supplementary analyses of Black respondents are reported in the online supplement and briefly summarized below. For more information on NSAHS data collection, see Turner, Brown, and Hale (2017).

Measurement

Neighborhood racial composition

We measure neighborhood racial composition with block group-level data for the proportion of Black residents. These data represent 5-year averages taken from the American Community Survey that overlap the NSAHS study period (2010–2014). Higher scores reflect larger proportions of Black residents in a respondent’s block group.

Perceived neighborhood danger

We measure perceptions of neighborhood danger with the following three items: (1) perceived likelihood that someone’s house, apartment, or car would be broken into (1 = very likely, 4 = very unlikely); (2) perceived likelihood someone would be physically threatened if walking alone on the street (1 = very likely, 4 = very unlikely); and (3) perceptions of the neighborhood being “pretty safe” (1 = strongly agree, 4 = strongly disagree). All three items were scaled such that higher numbers reflect a greater sense of danger. Similar items have been used in past studies to predict mental and physical health outcomes. For instance, Hill et al. (2016) confirmed that a comparable measure of perceived neighborhood safety predicted behavioral health outcomes across six distinct national contexts.

Educational attainment

Our key SES indicator is education. Respondents were asked to report their highest level of education in years and degrees completed. Original categories for completed degrees include less than high school, high school/GED, some college, college graduate, and postgraduate. We collapsed these categories into two groups of college educated (i.e., college graduate or higher) and non-college educated (i.e., high school, high school/GED, some college).² A review of the literature suggests that the college/non-college binary captures meaningful health disparities above and beyond years of education (Lawrence 2017; Liu et al. 2011, 2013). Proponents of this measure argue that college degrees are distinct from years of education, as they signal acquired habits and skills that confer unique socioeconomic benefits (Lawrence 2017; Liu et al. 2011, 2013). Moreover, midlife mortality is rising among Whites without college degrees (Case and Deaton 2015, 2017). This suggests college degrees are meaningful status distinctions among Whites.

Subjective social status

We measure respondents’ status appraisals using a single survey measure of subjective social status (SSS). Respondents were shown an image of a 10-rung ladder and then told, “The steps on the ladder stand for the 10 possible steps in your life. Level 9 stands for the best possible way of life for you and the first step stands for the worst possible way of life for you.” Respondents were then asked to report the number of the step corresponding with where they see themselves now. Scores are coded such that higher numbers reflect higher SSS.

Researchers have conceptualized SSS as a reflection of one’s perceived standing in broader socioeconomic hierarchies (Singh-Manoux, Adler, and Marmot 2003; Adler et al. 1994). Sociologists have also conceptualized SSS as a partial component of perceived “social power” or one’s perceived ability to impact society and effect change in one’s environments (Bierman, Lee, and Schieman 2018). Studies consistently find that people who report higher SSS also report better health outcomes cross-sectionally and over time, regardless of objective status indicators like education, occupation, or income (Hoebel and Lampert 2020).

Psychophysiological distress

Our focal outcome is psychophysiological distress, measured as a second-order latent construct consisting of first-order latent variables for somatization, anxiety, and anger. Somatization is indexed by the following past-month symptoms: (1) You could not “get going”; (2) You did not feel like eating; (3) You had trouble keeping your mind on what you were doing; and (4) You felt that everything you did was an effort. Response options range from “not at all” (=1) to “almost all the time” (=4). Anxiety is indexed by the following past-month symptoms: (1) I felt anxious; (2) I felt worried over possible misfortunes; (3) I felt tense; and (4) I felt nervous. Response options range from “not at all” (=1) to “very much” (=4). Anger is indexed by the following six items: (1) I feel angry; (2) When I get angry, I stay angry; (3) I yell at people; (4) I feel like I am boiling inside; (5) I lose my temper; and (6) I get into fights and arguments. Response options range from “not like me at all” (=1) to “very much like me” (=5). Similar items have been used in prior studies. For example, Ross and Mirowsky’s (2009) study of neighborhood effects on mental health used comparable items to construct a higher-order latent variable they labeled “agitation,” which comprised lower-order latent variables of anxiety, anger, and “malaise” (i.e., somatization).

Covariates. Multivariable analyses include respondent-level covariates for age (in years), gender (1 = female, 0 = male), marital status (1 = married, 0 = not married), employment status (1 = employed, 0 = non-employed), and household income (ordinal; 0 = no income, 15 = $135 k or more). To better isolate variance in perceived neighborhood danger that is attributable to neighborhood racial compositions, we also control for block group-level total crime rates. Crime data are retrieved from the FBI’s Uniform Crime Reports averaged over 2013 to 2019 (ArcGIS 2020). Total crime rates include personal crimes (e.g., murder) and property crimes (e.g., burglary). Rates reflect an index value for the average of US law enforcement jurisdictions. For example, a block group that scores 1.55 has a total crime rate that is 55% higher than the national average of law enforcement jurisdictions.

Analytic strategies

We test our study hypotheses using multigroup structural equation modeling (SEM) techniques with latent variables (Figure 2). Our main SEM is tested separately for Whites with and without college educations (not depicted). The full SEM comprises distinct measurement and path components. The measurement model refers to the associations among latent variables and their respective indicators. We correlate the errors of three anger indicators to account for measurement artifacts, as these items gauge outward expressions (vs. feelings) of anger. Full confirmatory factor analyses of the measurement model are reported in the online supplement (Table 1A–B).

Figure 2.

Structural equation model with latent variables: NSAHS, White subsample (N = 625). Notes: Model is tested separately for Whites with and without college educations. Chi-square (df) = 753.32 (479). CFI = .94. 1-RMSEA = .96. BIC_k = −2330.36.

Table 1.

Weighted Descriptive Statistics: NSAHS, White Subsample (N = 625)

	Non-college educated		College-educated
	(n = 316)		(n = 309)
Neighborhood context
Proportion Black residents	.22	(.21)	.14	(.15)***
Total crime rate	1.53	(.69)	1.18	(.79)**
Perceived neighborhood danger (range, 1–4)
Likelihood of physical threat	1.59	(.74)	1.44	(.65)*
Likelihood of break-ins	2.25	(.98)	2.12	(.90)
Neighborhood is “pretty safe” (reverse)	1.76	(.62)	1.61	(.59)**
Subjective social status (range, 1–10)	7.26	(1.66)	7.94	(1.40)***
Somatization (range, 1–4)
Cannot “get going”	1.88	(.84)	1.64	(.72)**
Does not feel like eating	1.51	(.70)	1.28	(.54)***
Trouble focusing	2.00	(.82)	1.86	(.66)*
Everything an effort	2.10	(.96)	1.77	(.75)***
Anxiety (range, 1–4)
Feels anxious	2.11	(.99)	2.19	(.87)
Worries over misfortunes	1.91	(.99)	1.82	(.87)
Feels tense	2.12	(.97)	2.22	(.87)
Feels nervous	1.96	(.94)	1.93	(.86)
Anger (range, 1–5)
Feels angry	2.36	(1.07)	2.18	(1.04)*
Gets angry, stays angry	2.20	(1.06)	1.97	(.95)**
Boiling inside	2.12	(1.03)	1.87	(.95)**
Yells at people	2.21	(1.20)	1.87	(1.00)***
Loses temper	2.28	(1.05)	2.08	(.97)*
Gets into fights/arguments	1.69	(.88)	1.56	(.79)*
Covariates
Age (in years)	46.31	(11.61)	42.84	(11.78)**
Female (vs. male)	.49	(.50)	.52	(.50)
Married (vs. not)	.61	(.49)	.71	(.45)**
Employed (vs. not)	.72	(.45)	.85	(.35)***
Household income (range, 0–15)	8.58	(3.57)	11.08	(3.33)***

Notes: Means/proportions are reported with standard deviations in parentheses.

^* p < .05, ^**p < .01, ^***p < .001 significant between-group difference (two-tailed).

The path model refers to the associations among latent variables and their exogenous predictors. Covariates are allowed to correlate with neighborhood racial composition and predict all endogenous variables in the path model (not depicted). Results for the path model are reported in two tables. Table 2 reports direct path (i.e., regression) coefficients. Table 3 reports decompositions of indirect paths between the proportion of Black residents at the neighborhood level and distress, via perceived danger and SSS. Sobel (1982) tests are used to identify significant indirect pathways.

Table 2.

Results from the Multigroup Path Model: NSAHS, White Subsample (N = 625)

	Not College-Educated (n = 316)						College-Educated (n = 309)
	Perceived danger		SSS		Distress		Perceived danger		SSS		Distress
Neighborhood Black residents	.212	(.094)*	.168	(.040)***	−.074	(.073)	.333	(.100)**	−.040	(.062)	.024	(.068)
Perceived neighborhood danger	−		−		.262	(.074)***	−		−		.197	(.063)**
Subjective social status (SSS)	−		−		−.568	(.088)***	−		−		−.421	(.100)***
Age	−.002	(.001)*	.002	(.001)**	−.001	(.001)	.000	(.001)	.000	(.001)	−.001	(.001)
Female	−.003	(.022)	−.002	(.022)	.055	(.025)*	.041	(.024)	.060	(.020)**	.018	(.018)
Married	−.020	(.034)	.093	(.019)***	.044	(.031)	.004	(.035)	.026	(.020)	.057	(.020)**
Employed	−.031	(.032)	.069	(.027)*	−.018	(.026)	−.043	(.040)	.008	(.025)	−.019	(.026)
Household income	−.005	(.004)	.008	(.003)*	−.006	(.004)	−.004	(.004)	.015	(.005)**	−.003	(.004)
Neighborhood crime	.047	(.033)	−.017	(.014)	.020	(.018)	.054	(.019)**	.002	(.008)	−.006	(.009)
Intercept	.265	(.115)*	.419	(.041)***	.691	(.080)***	.082	(.076)	.528	(.059)***	.574	(.071)***
R-squared	.160	(.075)*	.201	(.038)***	.446	(.056)***	.211	(.087)*	.169	(.057)**	.341	(.081)***

Notes: Weighted unstandardized regression coefficients are presented with robust standard errors clustered by block group in parentheses. Endogenous variables are rescaled to range from 0 to 1.

^* p < .05, ^**p < .01, ^***p < .001 (two-tailed).

Table 3.

Path Decompositions: NSAHS, White Subsample (N = 625)

	Not college-educated		College-educated
Black residents → perceived danger → distress → somatization	.056	(.025)*	.065	(.028)*
Black residents → perceived danger → distress → anxiety	.073	(.034)*	.104	(.043)*
Black residents → perceived danger → distress → anger	.042	(.019)*	.055	(.024)*
Black residents → SSS → distress → somatization	-.095	(.032)**	.017	(.027)
Black residents → SSS → distress → anxiety	-.125	(.038)**	.027	(.042)
Black residents → SSS → distress → anger	-.072	(.024)**	.014	(.022)

Notes: Weighted unstandardized coefficients are presented with robust standard errors clustered by block group in parentheses. SSS = subjective social status.

^* p < .05, ^**p < .01 (two-tailed).

The full SEM is tested in Mplus 7. Estimates are derived using full information maximum likelihood (FIML) procedures, with probability weights and robust standard errors clustered by block group. FIML is chosen because it is superior to listwise deletion and multiple imputation for recovering missing observations in SEMs (Enders and Bandalos 2001). Findings generated with FIML are comparable to the results after listwise deletion. The full model exhibits a good fit to the data structure. The CFI and 1-RMSEA score above the minimum accepted threshold of .90, or at .94 and .96, respectively (Weston and Gore 2006). The BIC_k score is −2330.36, indicating that the estimated model is superior to a fully saturated model (Raftery 1995).

Results

Characteristics of the NSAHS White subsample

Table 1 reports weighted descriptive statistics of study variables in two separate columns for White respondents with and without college educations. In short, compared with their college-educated peers, non-college-educated Whites tend to live in block groups with more Black residents and higher crime rates. Non-college-educated Whites also report greater perceptions of neighborhood danger, heightened symptoms of anger and somatization. They are also less likely to be married or employed, tend to be a few years older, and report lower average household incomes.

Test of the vigilance hypothesis

Path model results are presented in Tables 2 and 3. To facilitate model convergence, all endogenous indicators have been rescaled to range from 0 to 1. The results in Table 2 support the vigilance hypothesis. For college-educated Whites, a one-unit increase in the proportion of Black residents at the block group level predicts a .33-unit increase in perceived neighborhood danger, conditional on covariates (b = .33; p < .01). For non-college educated Whites, we find that perceived danger is expected to increase by .21 units for every one-unit increase in Black residents (b = .21; p < .05).

When multiplied by 100, these coefficients can be interpreted as the expected percentage-increase in perceived danger between Whites who live in all-White versus all-Black block groups. Among non-college-educated Whites, for example, this would reflect a 21% increase in perceived danger from residing in all-Black (vs. all-White) neighborhoods, whereas college-educated Whites are expected to report a similar increase of 33%. Importantly, these patterns persist net of block group crime rates, which also predict perceived danger for college-educated Whites.

An increased sense of neighborhood danger, in turn, predicts greater symptoms of distress. For college-educated Whites, a one-unit increase in perceived danger predicts a .19-unit increase in distress (b = .19; p < .01). For non-college-educated Whites, we expect to observe a .26-unit increase in distress for every one-unit increase in perceived danger (b = .26; p < .001). Moreover, we observe no direct associations between block group proportions of Black residents and distress for either group, suggesting there are only indirect paths via perceived danger.

Table 3 confirms that block group proportions of Black neighbors indirectly predict greater symptoms of distress via perceived danger. For non-college-educated Whites, a one-unit increase in the proportion of Black residents ultimately predicts increases in somatization, anxiety, and anger of .05, .07, and .04 units, respectively, on a zero-to-one scale (p < .05). Multiplied by 100, these coefficients reflect 4–7% increases in distress symptoms between non-college educated Whites in all-Black (vs. all-White) block groups. For college-educated Whites, we observe respective increases in somatization, anxiety, and anger of 7, 10, and 6% (p < .05).

Test of the comparison hypothesis

Tables 2 and 3 also provide support for the downward comparison hypothesis among non-college-educated Whites. Table 2 shows that a one-unit increase in block group proportions of Black residents predicts a .17-unit increase in SSS for non-college-educated Whites only (b = .17; p < .001). This coefficient reflects an expected 17% increase in SSS between non-college-educated Whites who live in all-Black (vs. all-White) block groups.

Table 3 confirms that higher SSS ultimately offsets distress for non-college-educated Whites who live in block groups with more Black residents. Looking at the first column of Table 3, we find that symptoms of somatization, anxiety, and anger are ultimately expected to decrease by .09, .12, and .07 units, on a zero to one scale, between non-college educated Whites who live in all-Black (vs. all-White) block groups. Considered together with the increases in distress via perceived danger, higher SSS entirely offsets distress for non-college-educated Whites residing in neighborhoods with more Black neighbors.

Supplementary analyses

We conducted sensitivity analyses reported in the online supplement. First, we tested for non-linear patterns between Black neighbors and perceived danger and SSS (Supplemental Table 2). Although findings were consistent with linear associations, we found some evidence to suggest that patterns were especially pronounced for Whites who lived in block groups that were 30% Black or higher.

Second, we tested additional controls (Supplementary Tables 3–6). At the resident level, we added controls for age-squared, children in the home (count), and a retrospective measure of childhood financial status (ordinal; 0 = family could not afford food/clothing/shelter, 4 = family could easily afford food/clothing/shelter with lots of extras). At the block group level, we also included additional 5-year ACS estimates (2010–2014) of residential stability (proportion of residents in the same house for at least one year), and proportion of college educated residents. In short, although there are slight attenuations in some path coefficients, our main findings are robust to these controls.

Third, we tested a multigroup SEM stratified by the median score of the first principal component for education, household income, and childhood financial status (eigenvalue = 1.45; shared variane = .49). Findings in this model reveal similar patterns (Supplementary Table 6).

Finally, we tested the model in Figure 1 among Black NSAHS participants (Supplementary Table 7). Briefly, neighborhood racial composition is not associated with perceived danger or SSS for Black respondents regardless of education level. For non-college-educated Black residents, however, the proportion of Black residents predicts fewer symptoms of distress (b = −.078; p < .05). These patterns also suggest that Whites are uniquely sensitive to neighborhood racial compositions, specifically in terms of their perceived safety and status.³

Discussion

Researchers have long documented how anti-Black racism harms the health of Blacks (Williams 2012; Williams and Mohammed 2013). But little attention has been given to whether and how anti-Black racism harms Whites. To fully grasp the toxicity of racialized social systems in the United States, it is important for researchers to examine how racism harms all members of society (Malat, Mayorga-Gallo, and Williams 2018; McGhee 2021; Metzl 2019). We advance current literature by showing how the positioning of Whites within racialized hierarchies can affect their health by exposing them to unique psychosocial stressors and resources stemming from neighborhood contexts.

Our analyses revealed two key findings. First, Whites residing in areas with more Black residents tended to report a greater sense of danger in their neighborhoods and, ultimately, more symptoms of distress. These findings persisted even after adjusting for neighborhood crime rates and socioeconomic compositions, suggesting that the mere presence of Black neighbors generated fear and distress for Whites. We have argued that these patterns stem from longstanding Black-White relations in the United States. Indeed, conceptions of White identity have been inextricably linked to the denigration of Blackness as a mark of inferiority, danger, and criminality (Bonilla-Silva 2003; Yancey 2008). In our study, we identified fears of Blackness as manifestations of systemic anti-Black racism, which can ironically harm Whites by triggering distress in the presence of Blacks.

Our second key finding is that lower-status Whites derived unique mental health benefits from living near more Black neighbors. To be specific, low-SES Whites who lived in neighborhoods with greater proportions of Black residents also tended to report higher SSS, which entirely offset distress from perceptions of danger. No similar patterns emerged among higher-SES Whites, who only reported increased perceptions of danger and distress around more Black neighbors. Drawing from historical and contemporary records of Black-White relations in the United States, we hypothesized that low-SES Whites may derive greater psychosocial benefits from living near Black neighbors because they have fewer resources to draw from when appraising their relative social standing. Thus, the value of Whiteness for low-status Whites appears to be tied to compensatory psychological resources from their perceived superiority to marginalized Black reference groups (Du Bois 1935; Mills 1997; Roediger 1991; Yancey 2008).

Our findings have broad implications. First, despite benefitting from centuries of structural racism in the United States, Whites tend to exhibit more psychiatric disorders than Blacks (Barnes and Bates 2017). This general pattern is commonly referred to as the “Black-White mental health paradox” (Louie et al. 2022; Louie and Wheaton 2018). Most research in this area attempts to explain the paradox by highlighting unique sources of resilience among Blacks, such as high self-esteem and religious involvement (Louie et al. 2022; Louie and Wheaton 2018; Mouzon 2017; Upenieks, Louie, and Hill 2023). Our study advances this line of work by considering unique sources of vulnerability among Whites, particularly anti-Black vigilance (Malat, Mayorga-Gallo, and Williams 2018; Schnittker and Do 2020). To be sure, our findings cannot resolve whether anti-Black vigilance is responsible for higher rates of psychiatric disorder among Whites, but future research should consider this.

The points raised thus far support Pattillo’s (2021) emerging framework of Black Advantage Vision (BAV). BAV inverts common approaches to racism research by centering the “strengths, worth, resilience, care, and accomplishments” of Blacks (Pattillo 2021, 79), inviting scholars to consider the advantages of Blackness and disadvantages of Whiteness (see also Malat, Mayorga-Gallo, and Williams 2018). From a BAV standpoint, our study is consistent with research showing that Whites typically prefer living in racially homogenous communities (e.g., Krysan and Farley 2002), and often report feeling less connected with their communities than Blacks (Taylor et al. 2022). Our study also supports works finding that Blacks typically have better subjective well-being than Whites (Graham and Pinto 2019), in part due to positive Afrocentric worldviews that emphasize communality and spirituality (Louie et al. 2022; Neblett and Carter 2012; DeAngelis, Upenieks, and Louie 2023). Our study reiterates Pattillo’s (2021) central insight that researchers should stop treating Whites as the yardstick by which to measure Black progress. White communities are fraught with problems that warrant their own study (Zuberi and Bonilla-Silva 2008).

On a related note, Case and Deaton (2015, 2017) find that midlife mortality rates have been rising for lower-status Whites over the last twenty years. This appears to be mostly due to increases in suicide, drug and alcohol poisonings, and alcohol-related liver disease (a.k.a., “deaths of despair”). Several newer studies have advanced the “racial status threat” hypothesis to explain these trends (Siddiqi et al. 2019; Rambotti 2022). Accordingly, “as marginalized groups increase in number, whites are more likely to worry that their status – their superior position in society – is declining” (Siddiqi et al. 2019, 10). Our study qualifies this hypothesis. Indeed, low-status Whites in our study appeared to derive perceptions of higher status and related mental health benefits from living near more Black residents. Our study points, instead, to racialized fears of crime and victimization as more specific mechanisms of distress for Whites in an increasingly diverse and globalizing society (Duxbury 2021b, 2021a, 2023). Future research should determine whether anti-Black vigilance is somehow contributing to increasing stress and “despair” among lower-status Whites.

Our findings also speak to an emerging body of research into the fraught relations within multiracial settings. For instance, in a 2014 book, Behind the White Picket Fence, Sarah Mayorga-Gallo found that White residents of a racially integrated neighborhood in Durham, North Carolina publicly lauded the benefits of racial diversity, yet avoided any meaningful contact with their Black neighbors. Other studies find that high-status Blacks who occupy prestigious neighborhoods, universities, and occupations tend to report more discrimination and stress than their White and lower-status Black peers (DeAngelis 2022; DeAngelis, Hargrove, and Hummer 2022). We advance this work by providing new evidence that anti-Black vigilance among Whites likely contributes to these strained relations (Anderson 2015; Feagin and Sikes 1994; Lucas 2008; Yancey 2008).

Other similar processes related to anti-Black vigilance may also generate stress for Whites in multiracial workplaces or classrooms. For example, evidence suggests that Whites in professional settings sometimes experience stress and anxiety from concerns over appearing racist in front of Black colleagues (Plant and Devine 2003; Trawalter et al. 2012). Although these processes may be less stressful than fears of crime and victimization, chronic self-regulation could also generate distress over long periods. Future studies should inquire into how often Whites enact the following behaviors at work or school: (1) regularly prepare for possible encounters with Black colleagues; (2) feel they have to be careful about what they say in front of Black colleagues to avoid being taken the wrong way; and (3) try to avoid certain social situations that involve Black colleagues. These items may tap into similar feelings of vigilance related to encountering Blacks in social spaces (e.g., Hicken et al. 2013).

A research program that incorporates our findings and recommendations would expand our understanding of how anti-Black racism functions as a multifaceted system that affects population health (Reskin 2012). For instance, new efforts are underway to measure manifestations of racism and their health effects across multiple geographical, political, and cultural contexts (Brown and Homan 2022). Our study adds a critical and largely overlooked dimension to this emerging research program: racialized emotions, particularly fear and distress from Whites (see also Bonilla-Silva 2019; Duxbury 2021b, 2023). Future research in this area should consider how to systematically measure Whites’ fears of marginalized groups across different spatial-temporal scales. For instance, an emerging area of research uses aggregated scores from the Implicit Association Test to predict county- and state-level racial disparities (Leitner et al. 2016; Payne, Vuletich, and Brown-Iannuzzi 2019). Measures like these can help researchers better identify hotbeds of racism in the United States. They may also reveal contexts where racialized health inequities are most severe.

At the neighborhood level, scholars should also examine whether our findings generalize to other major cities like Washington, DC, Boston, or Chicago (e.g., Bader and Krysan 2015). For example, one study in the Boston area found that White residents of multiethnic neighborhoods also reported fearing for their safety (Walton 2021). Yet evidence from Washington, DC, indicated that White residents in multiracial neighborhoods were no less satisfied about their living conditions than their Black peers (Bader 2022). Clearly, more research is needed to unpack how Whites from different areas and socioeconomic backgrounds interpret their neighborhood spaces, and how such appraisal processes ultimately affect their health and interactions with neighbors.

Limitations

This study is not without limitations. First, as others have noted (DeAngelis 2022; Turner, Brown, and Hale 2017), the sociohistorical contexts of Nashville are unique and may not generalize to other areas. Additional research is needed to understand whether the relationships we uncovered among White Nashvillians generalize to other areas in the United States, or other stigmatized groups like Latino and Muslim Americans. For example, a study of Miami-Dade County found that a higher percentage of Latino neighbors predicted greater fears of crime among White residents (Eitle and Taylor 2008). Their work reiterates that although similar patterns may surface among Whites in other cities, attention to local histories and immigration patterns is needed.

Second, the NSAHS is cross-sectional, and thus definitive causal inferences cannot be made. We also cannot account for neighborhood selection processes, including how long respondents lived in their neighborhoods or their reasons for moving in. There was also a one-year lag between initial screening and interviewing for the NSAHS, during which time some participants may have moved from their initial block group, adding measurement error to our estimates. Future research could use longitudinal data to better account for potential endogeneity. That said, although some evidence suggests those with poor mental health also perceive situations to be more stressful, the bulk of evidence indicates that social stressors cause mental health rather than vice versa (Brown et al. 2000). Moreover, assuming Whites who are most fearful of Black neighbors will avoid racially integrated areas, our estimates of distress are likely conservative.

Finally, our models also assumed that the associations between neighborhood racial compositions, status appraisals, and mental health form linear patterns among Whites. Yet our study hypotheses implied there could be discernible groups of Whites who perform a variety of discrete racialized social comparisons. Future research could test these alternative assumptions using latent class analysis or in-depth qualitative interviewing, either of which could help to reveal distinct groups of Whites who draw from similar contextual influences to inform social comparisons.

Conclusion

Our study revealed contexts in which anti-Black racism can benefit or harm the health of Whites. In our study, Whites who lived near more Black neighbors felt their neighborhoods were more dangerous regardless of actual crime rates. These perceptions, in turn, were associated with higher distress. For low-SES Whites, however, living near more Black neighbors also conferred a sense of higher status, which buffered distress. As scholars push to advance our understanding of anti-Black racism and White supremacy, distinctions among White groups should not be overlooked. Importantly, future studies should determine whether different White groups have other context-dependent coping resources that help them alleviate racialized fears in ways that do not entail the denigration of Black Americans as a perceived inferior reference group.

This research uses data from the Nashville Stress and Health Study, a project led by the late R. Jay Turner and funded by the National Institute on Aging (R01AG034067). Reed DeAngelis received support from the Duke Aging Center Postdoctoral Research Training Grant (NIA T32-AG000029), as well as the Population Research Infrastructure (P2C-HD050924) and Biosocial Training (T32-HD091058) programs awarded to the Carolina Population Center at the University of North Carolina at Chapel Hill by the Eunice Kennedy Shriver National Institute of Child Health and Human Development. The authors also thank Kyle Crowder for helpful feedback on earlier drafts of our study. Any errors or omissions are the fault of the authors and not of their reviewers.

About the author

Patricia Louie is an assistant professor of sociology at the University of Washington. Her research focuses on the social determinants of health, with a focus on the mechanisms that underlie racial disparities in mental and physical health. Her recent research has been published in Journal of Health and Social Behavior, JAMA Network Open, Society and Mental Health, Social Science and Medicine, and Social Psychology Quarterly.

Reed DeAngelis is a postdoctoral researcher at the Duke University School of Medicine. He is a biodemographer and medical sociologist who researches social-environmental inequities in health and aging. Some of his recent studies are published in Demography, Social Forces, Journal of Health and Social Behavior, and Journal of Racial and Ethnic Health Disparities.

Data Availability

The data underlying this article cannot be shared publicly for the privacy of individuals that participated in the study.