Perceived Neighborhood Disorder, Self-Esteem, and the Moderating Role of Religion

Abstract

A growing body of work links neighborhood conditions –and particularly perceived neighborhood disorder—with diverse aspects of psychosocial functioning, including self-esteem or the global moral self-worth of the individual. Our work augments this literature by investigating the possible roles of (a) organizational religiosity (i.e., religious attendance, religious support), (b) non-organizational religiosity (i.e., prayer and religious coping practices), and (c) the sense of divine control as potential stress in mitigating the deleterious effects of neighborhood disorder on self-esteem. Data from the Nashville Stress and Health Study (NSAHS, 2011–2014) are used to test a series of hypotheses regarding the possible stress-buffering effects of multiple religious domains. Findings from multivariable regression models indicate that: (a) perceived neighborhood disorder is inversely associated with self-esteem; (b) non-organizational religiosity and the sense of divine control each mitigate this pattern; and, interestingly, (c) organizational religiosity does not buffer the association between neighborhood disorder and self-esteem. Several study limitations, as well as a number of promising directions for future research, are identified.

INTRODUCTION

Self-esteem refers to assessments of the self as an object, and is typically indicated by perceptions of worth and competence (Mruk 2006; Rosenberg 1981). Scholars have investigated patterns, correlates, and sequelae of self-esteem for several reasons. First, positive self-image is often regarded as one facet of psychological well-being, a desirable end in itself (e.g., Bradshaw and Kent 2018). Second, self-esteem has been viewed as an important motivator of human cognition and behavior, as people often process information in ways that conserve or enhance positive self-regard (Crocker and Park 2004; Sedikides and Gregg 2008). Third, self-esteem is also an important psychosocial resource that contributes to better mental health and greater life satisfaction by promoting successful coping with chronic and acute stressors (Orth and Robbins 2013; Taylor and Stanton 2007). Finally, although some observers argue that the role of self-esteem may be exaggerated (e.g., Baumeister et al. 2003), self-esteem may also predict a host of other outcomes, including human capital and labor market outcomes, financial stability, health and health behavior, criminal and delinquent conduct, and marriage (for a review, see Maclean and Hill 2015).

In light of these findings, social scientists have identified factors that promote or undermine self-esteem, such as material success and status attainment, racial/ethnic minority status and identity, marital and intimate relationships, social ties, health, and well-being (Hughes and Demo 1989; Rosenberg and Pearlin 1978; Waite, Luo, and Lewin 2009). One productive line of inquiry has linked aspects of residential context, especially the elements of neighborhood economic structure and perceptions of disorder, with self-esteem, in both cross-sectional and longitudinal data (Behnke et al. 2011; Haney 2007; Hill et al. 2013; Kim and Pai 2022; McMahon, Felix, and Nagarajan 2011; Pederson et al. 2022; Prelow, Weaver, and Swenson 2006; Romero et al. 2020). In this study, we focus on neighborhood disorder. This construct, described more fully below, refers to a range of adverse social (e.g., crime) and physical (e.g., abandoned buildings) conditions that indicate the breakdown of social control in the neighborhood. In contrast to other indicators of neighborhood conditions, neighborhood disorder centers on the perceptions and lived experiences of residents (Hill and Maimon 2013).

Given that the association between neighborhood context and self-esteem is generally understudied, it is perhaps unsurprising that little is known about potential subgroup variations. Although there is some evidence to suggest that the effects of neighborhood context on self-esteem may vary by gender, race, social network size, familism (shared family activities), social support, and marital status (Behnke et al. 2011; Kim and Pai 2022; McMahon, Felix, and Nagarajan 2011; Pederson et al. 2022; Prelow, Weaver, and Swenson 2006; Romero et al. 2020), it is unclear whether different dimensions of religious involvement may moderate the deleterious association between perceived neighborhood disorder and self-esteem. There are several reasons why this issue is worthy of investigation. First, although a long tradition of work has argued that aspects of religious belief may erode self-esteem (e.g., Ellis 1962; Watters 1992), most contemporary empirical studies suggest that religiosity is conducive to a more positive self-concept (Bradshaw and Kent 2018; Ellison, Schieman, and Bradshaw 2014; Krause 2003, 2009; Krause and Hayward 2014; Schieman et al. 2005, 2017). Second, a wealth of theory and research has linked facets of religiosity with successful coping under stressful conditions (Ellison and Henderson 2011; Pargament et al. 2000; Smith et al. 2003). Indeed, several studies have specifically investigated the role of various religious and spiritual domains in adjustment to the strain of living in dilapidated or deteriorating neighborhoods (e.g., Krause 1998, Krause et al. 2017).

This study contributes to previous work in two key respects. First, given that previous studies of neighborhood context and self-esteem have mostly focused on children and adolescents, we attempt to replicate the inverse association between neighborhood disorder and self-esteem in adulthood. Second, because no research to date has investigated the role of religious factors in moderating the links between neighborhood context and self-esteem, we explore whether three facets of religion–organizational involvement, non-organizational religious involvement, and perceived divine control—might mitigate the effects of neighborhood disorder. We examine these issues using data from the Nashville Stress and Health Study (NSAHS), a probability sample of African American and non-Hispanic White adult residents of Davidson County, TN (n = 1,252). After presenting results from regression models, we discuss the implications of our findings for future work on the complex linkages among neighborhood disorder (and other stressors), religious factors, and health and well-being.

BACKGROUND

Perceived Neighborhood Disorder and Self-Esteem

Research on health, well-being, and psychosocial functioning has incorporated aspects of neighborhood conditions for some time. Such conditions are often gauged in terms of the characteristics of Census tracts, interviewer assessments (in studies using data from face-to-face interviews), and other means that do not always rely on, or capture, the self-reports of residents regarding their own understandings of these residential conditions (for a review, see Hill and Maimon 2013). The concept of perceived neighborhood disorder has been characterized by Ross and Mirowsky (1999:413) as “...visible cues indicating a lack of order and social control in the community.” Perceived neighborhood disorder is unique because it taps the lived experiences of residents with regard to areas that they define as their neighborhoods, as opposed to Census tracts or other standardized geographic units.

Ross and Mirowsky (1999), and others who have followed their line of reasoning, identify several types of cues that index the level of disorder within a neighborhood. Social disorder is reflected by such indicators as conflict among neighbors, the presence of people drinking, taking drugs, and panhandling, and other cues that may result in a sense of threat and danger. Physical disorder is reflected by the overall appearance of the neighborhood, including dirtiness and noisiness, buildings that are abandoned or in disrepair, and overgrown lots, graffiti and other signs of neglect. High levels of social and physical disorder may indicate a breakdown in the conventional mechanisms of social control (i.e., shared norms of maintenance and civility, collective efficacy, etc.) that typically govern orderly neighborhoods.

Over the past 25 years or so, empirical studies have linked perceived neighborhood disorder with an array of psychosocial and behavioral outcomes, including, for example, mistrust, social isolation, perceived powerlessness, religious and spiritual struggles, heavy drinking, sleep disturbance, psychological distress, and poor physical health (Blair et al. 2014; Carbone 2020; Geis and Ross 1998; Ross and Jang 2000; Ross et al. 2001; Hill, Burdette, and Hale 2009; Hill, Ross, and Angel 2005; Hill et al. 2023 Intravia et al. 2016; Johnson, Billings, and Hale 2018; Kim and Conley 2011; O’Brien, Farrell, and Welsh 2019; Ross and Mirowsky 2009). Studies have also reported inverse associations between perceived neighborhood disorder and self-esteem (Behnke et al. 2011; Haney 2007; Hill et al. 2013; McMahon, Felix, and Nagarajan 2011; Pederson et al. 2022; Prelow, Weaver, and Swenson 2006).

Why might neighborhood disorder lead to diminished sense of self-worth? Prior theory and research have suggested several possible reasons. First, residents may be treated–or may perceive that they are treated–with disrespect and disregard by other persons (e.g., prospective employers, police, local officials, and even fellow residents) due to the problematic nature of their neighborhood (Behnke et al. 2011; Haney 2007; Kim and Pai 2022). To the extent that this occurs, as a result of what Rosenberg (1981) and others have termed “negative reflected appraisals,” they may find it difficult to view themselves as good, worthwhile individuals as deserving of respect as other persons. Second, residents may become aware of the disparities between the conditions in their disordered neighborhoods and the areas in which other citizens live. Such unflattering social comparisons may lead them to internalize a negative self-image (e.g., as unworthy, incompetent, etc.) (Haney 2007; Kim and Pai 2022). Third, to the extent that self-esteem may be rooted partly in perceptions of personal efficacy, as some scholars have theorized (e.g., Gecas and Schwalbe 1983), living in a neighborhood characterized by disorder and loss of social control may deepen feelings of powerlessness and inadequacy, especially since it may be difficult or impossible for residents to work individually or to mobilize collectively to reverse or ameliorate these conditions (Haney 2007; Kim and Pai 2022). Finally, perceived neighborhood disorder is associated with lower self-esteem because disorder tends to limit opportunities for social support (e.g., through processes related to stigma, fear, and mistrust) (Hill et al. 2013).

Taken together, these arguments lead to our first study hypothesis:

H1: Perceived neighborhood disorder will be inversely associated with self-esteem.

The Stress-Buffering Role of Religion

An important tradition of theoretical and case-based literature in psychology and allied fields claimed that religious belief undermines self-esteem and other facets of mental well-being. These arguments centered on the presumed roles of belief in an omnipotent deity and belief in human sinfulness. Such core theological tenets of orthodox Christianity have been thought to (a) foster negative self-appraisals directly and (b) to divert individuals’ energies from productive coping and other activities that could build perceptions of self-worth and personal competence (Ellis 1962; Watters 1992). In contrast to such work by clinicians and theorists, most contemporary empirical research has reported salutary associations between various aspects of religiosity and positive self-regard (Bradshaw and Kent 2018; Ellison 1993; Ellison et al. 2014; Krause 2003, 2009; Krause and Hayward 2014; Schieman et al. 2005, 2017). In what follows, we examine the possible roles of three distinct religious domains –organizational religiosity, non-organizational religiosity, and perceived divine control—in buffering or mitigating the deleterious association between neighborhood disorder and self-esteem.

Organizational Religiosity.

Religious congregations are network-driven institutions. Individuals and families are often recruited into congregations via pre-existing social networks and, after entry into the group, they may develop new, long-term network ties. Shared interests, values, and activities make religious groups fertile terrain for the cultivation of friendships and support networks (Ellison and George 1994; Krause 2008; Merino 2014). Indeed, a burgeoning literature demonstrates the role of congregational support systems, including both formal supports (e.g., church programs, pastoral counseling) and informal networks of exchange, in promoting positive psychosocial functioning and buffering the deleterious effects of chronic and acute stressors on aspects of well-being (Bradley et al. 2020; Krause 2008). Church members routinely exchange both instrumental aid (i.e., goods and services, information) and socio-emotional support (i.e., making others feel loved, valued, and cared for), in ways that foster well-being and successful adjustment to stressful events and conditions (Chatters et al. 2018; Krause 2008). It is reasonable to expect that the positive reflected appraisals from church-based support relationships may benefit self-esteem (e.g., Ellison 1993; but see Krause 2009), and may be particularly helpful for persons in disordered neighborhoods, which may make sustaining trusted friendships and support networks more challenging (Geis and Ross 1998; Ross et al. 2001).

In addition, prior theory and research suggest other potential reasons why organizational religious participation may mitigate the links between stressors and self-esteem. For example, regular attendance at services and other congregational participation may strengthen religious meaning systems (what Berger [1967] famously termed “plausibility structures”), which may help individuals to make sense of challenges and problems, thus reducing any impact of negative conditions on self-esteem. In addition, through engagement in congregational activities (e.g., social groups, religious education, charitable activities, etc.), some individuals may gain leadership opportunities and civic skills that may enhance perceived competence and confidence (Djupe and Gilbert 2006; Schwadel 2002), thereby potentially buffering self-esteem against the detrimental effects of stressors such as neighborhood disorder.

At the same time, there are reasons to be cautious in presuming an important stress-buffering role for organizational religiosity. For example, some congregations may promote theological systems that emphasize themes of human sinfulness and divine judgment, which may have the deleterious effects on self-esteem that some theorists have identified. In addition, recent research raises the possibility that the levels and psychosocial benefits of organizational religiosity may be contingent upon the degree to which individuals enjoy a secure attachment to God (Kent and Henderson 2017).

Nevertheless, on balance there is sufficient theoretical and empirical basis to support our second study hypothesis:

H2: Organizational religiosity will moderate (i.e., buffer) the inverse association between neighborhood disorder and self-esteem.

Non-Organizational Religiosity.

In addition to organizational religious activities, individuals may turn to prayer and faith when dealing with stressors such as neighborhood disorder and decay. Indeed, the use of religious coping practices may be especially common among persons dealing with material deprivation, marginality, and related problems. Often, such religious coping involves developing a dynamic, collaborative problem-solving strategy with God as a key component (Pargament 1997; Pargament et al. 2000). Such coping strategies are often helpful to persons facing financial strain, neighborhood disorder, and other conditions of social or economic strain or marginality (Krause 1998; Krause et al. 2017). Religion can also provide a meaning system that allows individuals to make sense of their world, derive hope from conditions of disadvantage, and reframe challenging conditions in terms that are less threatening to the sense of self (e.g., Vishkin et al. 2016).

Moreover, the practice of prayer may be especially conducive to well-being and elevated self-esteem. Briefly, individuals construct (perceived) personal relationships with God in much the same way in which they build social connections –i.e., through regular interactions experienced as conversations, informed by scriptural accounts of how God is believed to interact with the faithful. God is understood by most believers to be loving and caring (Bradshaw et al. 2010; Froese and Bader 2010). Thus, for many persons regular prayer may be a means to experience God as an ideal attachment figure, available for solace, guidance, and protection at any time (Luhrmann 2012; Pollner 1989; Schafer 2013; Sharp 2010). Persons who pray regularly to a loving God to whom they perceive a secure attachment tend to benefit with regard to self-esteem and other mental health outcomes, although such patterns are absent --and indeed, may be reversed-- for persons who pray to a colder, more distant deity (Bradshaw and Kent 2018; Ellison et al. 2014). An ongoing relationship with a perceived divine other –especially one that is experienced as loving and benevolent, which is the most common pattern among U.S. adults (Froese and Bader 2010)-- may be especially valuable in bolstering self-worth for persons living in communities with ever-present signs of disorder and decay, where suspicion and loss of trust among neighbors may be common (e.g., Ross et al. 2001). For people living in such communities, sustaining a personal relationship with God could be a particularly effective way to augment one’s sense of esteem and power while regularly navigating such precarious environments (e.g., Bandura 2003; Solomon, Greenberg, and Pyszczynski 1991).

Taken together, these strands of argument lead to our third study hypothesis:

H3: Non-organizational religiosity –prayer and turning to religion when coping—will moderate (i.e., buffer) the inverse association between neighborhood disorder and self-esteem.

Perceived Divine Control.

After a long period of neglect, scholars in the social psychology of religion are devoting fresh attention to the implications of religious cognitions –especially those involving perceptions of God-- for psychosocial functioning (Silton et al. 2014; Upenieks et al. 2022). One particularly important line of inquiry involves perceived divine control, which Schieman and colleagues (2006) characterize as “... the extent to which an individual perceives that God exerts a commanding authority over the course and direction of his or her life” (529). Thus, believers in divine control sense that they can reply on God for help and guidance, that God has already determined the way their lives will unfold, and that both positive and negative life outcomes are “part of God’s plan” (DeAngelis 2018; DeAngelis and Ellison 2017; Schieman et al. 2005, 2006, 2010; Upenieks et al. 2022).

Perceived divine control is a distinct religious cognitive schema with significant implications for stress and coping. Researchers in this field have conceptualized perceived divine control as a secondary coping strategy, specifically as a means of interpretive and vicarious control (DeAngelis 2018; DeAngelis and Ellison 2017). Briefly, the broad class of secondary control strategies entail modifying one’s expectations and self-perceptions to cope with circumstances beyond one’s control (Rothbaum, Weisz, and Snyder 1982). Interpretive control strategies specifically involve constructing benign rationalizations of distressing and potentially insurmountable conditions, while vicarious control strategies involve aligning oneself with powerful exogenous agents, such as a deity or a famed politician, to augment one’s own sense of esteem (Rothbaum et al. 1982: 20–24).

Perceiving that one’s life course is in the hands of God can afford vicarious esteem for persons facing challenges from economic marginality (DeAngelis 2018; Krause 2005)—or, in the current case, from the structural disadvantages of living in disordered neighborhoods. That is, believing that the social order (and one’s place in it) is shaped by an all-loving and all-powerful deity, who is in ultimate control of the future and who holds out the promise of eternal life for the faithful¸ could be an effective way to interpret one’s deprived material status in a more favorable way (e.g., DeAngelis and Ellison 2017). Indeed, according to Christian teaching, God –the most powerful entity in the universe—desires to love and care for the faithful, and to remain actively engaged in their lives. It has been suggested that such perceptions may well contribute to elevated self-worth via positive reflected appraisals (Ellison 1993; Krause 2009; Schieman et al. 2017).

Consistent with such arguments, several empirical studies have linked perceived divine control (or what Krause [2005] termed “God-mediated control”)—as well as belief in an engaged, loving and supportive deity more generally—with a number of favorable psychosocial outcomes, including enhanced self-esteem (Bradshaw et al. 2010; DeAngelis 2018; Schieman et al. 2005, 2006; Schieman et al. 2017; Silton et al. 2014; Upenieks et al. 2022). Importantly for the present study, perceived divine control has been shown to moderate the association between certain chronic stressors (e.g., frustrated aspirations, or goal-striving stress) and self-image (DeAngelis 2018; DeAngelis and Ellison 2018). Given this prior theory and evidence, it is reasonable to expect a similar pattern with regard to neighborhood disorder.

These strands of argument lead us to our final study hypothesis:

H4: Perceived divine control will moderate (i.e., buffer) the inverse association between neighborhood disorder and self-esteem.

Our study hypotheses are depicted below in Figure 1.

Figure 1.

Stress-Buffering Conceptual Model

METHODS

Data

We tested our hypotheses with data from Vanderbilt University’s Nashville Stress and Health Study (NSAHS), a probability sample of non-Hispanic Black and White adults living in Davidson County, Tennessee between 2011 and 2014. This is a cross-sectional project in which each respondent was surveyed only once during the data collection period. Researchers drew the sample using stratified multistage cluster sampling techniques to ensure a diverse representation of Black and White, working-age women and men. Investigators surveyed 1,252 adults living in a random sample of 199 block groups, which were chosen based on the percentage of African American residents according to 2010 Census data. The sampling frame consisted of 2,400 randomly selected households of which 2,065 were eventually contacted to participate in the study. Nearly 61 percent of the 2,065 households ultimately participated. Interviews were conducted either in the respondent’s home or on Vanderbilt University campus. Trained interviewers conducted the interviews and were matched to respondents based on race. The interviews were computer-assisted and lasted roughly three hours. Respondents were offered $50 to participate in the survey interview (for further information, see Turner et al. 2017).

Measures

Self-Esteem.

We measured self-esteem with the following six items: (1) “You feel that you have a number of good qualities,” (2) “You feel that you are a person of worth at least equal to others,” (3) “You are able to do things as well as most other people,” (4) “You take a positive attitude toward yourself,” (5) “On the whole you are satisfied with yourself,” and (6) “All in all, you are inclined to think you are a failure” (reverse-coded). Response categories ranged from “strongly disagree” = 1 to “strongly agree” = 5 (Rosenberg 1965). Scores were averaged and coded such that higher values reflect higher self-esteem (alpha = .81).

Neighborhood Disorder.

We used an index of perceived neighborhood disorder originated from Ross and Mirowsky`s (1999) scale. Assuming each item has equal weight, our index consisted of 13 items measuring the following four residential components: (1) physical disorder (e.g., “There are a lot of abandoned buildings in my neighborhood”); (2) physical order (e.g., “My neighborhood is clean”); (3) social disorder (e.g., “There is too much drug or alcohol use in my neighborhood”); and (4) social order (e.g., “In my neighborhood, people watch out for each other”). Response categories for all items ranged from “strongly disagree” = 1 to “strongly agree” = 4. The 13 items loaded onto a single factor with an eigenvalue of 6.10 and factor loadings ranging from .52 to .80. All items are coded such that higher scores reflect greater neighborhood disorder (alpha = .92).

Religious Involvement.

We measured organizational religious involvement with a mean index comprising the averages of three different items (alpha = .61). The first item asked respondents, “Which of the following best describes how often you attend services at a church/temple/synagogue/mosque?” Response categories ranged from 0 = “never” to 6 = “weekly or more.” The second item asked, “How often do you see, write, or talk on the telephone with members of your church (place of worship)?” Response categories ranged from 1 = “never” to 6 = “nearly every day.” The third and final item asked, “How often do people in your church (place of worship) help you out?” Response categories ranged from 1 = “never” to 4 = “very often.” Principal component analysis confirmed all items loaded onto a single factor with an eigenvalue of 1.86 and factor loadings ranging from .75 to .83. Preliminary analyses also confirmed each individual item produced comparable results in our multivariable analyses, which encouraged the use of an index for the sake of brevity.

Non-organizational religious involvement was measured via a mean index consisting of two averaged items (alpha = .80). The first item was frequency of prayer, which asked respondents, “About how often do you pray?” Response categories ranged from 1 = “never” to 6 = “several times a day.” The second item asked, “How often do you turn to your religious/spiritual beliefs to help you deal with your daily problems?” Response choices ranged from 1 = “never” to 5 = “always.” These two items also produced comparable results when tested individually.

We measured perceived divine control with the following four items (Schieman et al. 2005, 2006): (1) “I decide what to do without relying on God” (reverse-coded); (2) “When good or bad things happen, I see it as part of God’s plan for me”; (3) “God has decided what my life shall be”; and (4) “I depend on God for help and guidance.” Response categories ranged from 1 = “strongly disagree” to 4 = “strongly agree.” We averaged the items to create a mean index of perceived divine control (alpha = .84).

Stressors.

Models also account for recent life events and financial strain, both of which could confound the unique effects of neighborhood disorder on self-esteem. First, we included a check-list inventory of 32 major life events that could have happened to the respondent in the past year. These events included a serious injury or illness, an unwanted pregnancy, the death of a loved one, and the unexpected loss of a job, among others. We added together “yes” (=1) and “no” (=0) responses to create an additive index. Second, we included an 8-item mean index of financial strain (alpha =.77). The first set of questions asked whether respondents had difficulties meeting the following needs: (1) housing, (2) food, (3) transportation, (4) medical expenses, (5) other necessities, and (6) non-essential expenses. Response choices included “not at all difficult” = 1, “somewhat difficult” = 2, and “very difficult” = 3. The next question asked, “In general, how often do you worry about the total amount you (and your spouse/partner) owe in overall debt?” Responses ranged from “not at all” = 0 to “all of the time” = 4. The final question asked, “How concerned are you that you (and your spouse/partner) will never be able to pay off these debts?” Responses ranged from “not at all concerned” = 0 to “very concerned” = 4.

Socio-demographics.

Models also include covariates for age (in years), gender (1=female, 0=male), race (1=Black, 0=White), education (in years), marital status (1=married, 0=not married), employment status (1=employed fulltime, 0=part-time/unemployed), and household income (ordinal, 0=$5,000 or less, 15=$135,000 and above). We also controlled for an 8-item mean index of (non-religious) social support (alpha = .95). This index measured the extent to which respondents could count on their friends and family to make them feel worthwhile and appreciated, particularly during troubling times.

Analytic Strategies

Self-esteem is estimated using probability-weighted ordinary least squares (OLS) regression with Huber-White robust standard errors, clustered by block group, to adjust for homoscedasticity and permit generalizability to the broader population of Black and White adults in Nashville. To test our stress-buffering hypotheses, we include a series of (mean-centered) interaction terms between neighborhood disorder and religious measures. We also visually depict significant interaction terms as linear prediction graphs (Figures 2 and 3).

Figure 2.

Neighborhood Disorder × Non-organizational Religious Involvement on Self-Esteem.

Figure 3.

Neighborhood Disorder × Perceived Divine Control on Self- Esteem.

Finally, the following variables had missing values: self-esteem (n=5), neighborhood disorder (n=70), religious attendance (n=1), religious social support (n=1), perceived divine control (n=18), prayer (n=2), religious identity (n=9), household income (n=31), and social support (n=8). For all analyses, we replaced these missing values with 25 iterations of multiple imputation by chained equation (Stata’s “ice” command; Johnson and Young 2011).

RESULTS

Table 1 reports weighted descriptive statistics of study variables. On average, respondents report high levels of self-esteem and low to moderate levels of perceived neighborhood disorder. They also report moderate to high average levels of religiosity. For example, the average score on perceived divine control is 3.05. This suggests respondents tended to “agree” with the different items. Table 1 provides additional descriptive statistics.

Table 1.

Weighted Descriptive Statistics (NSAHS, 2011–2014)

	Range		Mean (%)	SD	Alpha
	Minimum	Maximum
Focal Measures
Self-esteem	1.16	5	4.59	0.54	.81
Neighborhood Disorder	1	3.62	1.74	0.52	.92
Religious Measures
Organizational religious involvement	.33	5.5	2.98	1.28	.61
Non-organizational religious involvement	1	5.5	4.21	1.31	.79
Perceived divine control	1	4	3.05	.76	.84
Stressors
Recent life events	0	15	2.14	2.17
Financial strain	.25	3.25	1.43	.53	.77
Socio-demographics
Age (in years)	22	69	46.04	11.52
Gender (1=female, 0=male)	0	1	(53)
Race (1=Black, 0=White)	0	1	(50)
Education (in years)	0	28	13.99	3.10
Marital status (1=married, 0=otherwise)	0	1	(45)
Employment (1= fulltime, 0=otherwise)	0	1	(60)
Household income (ordinal)	0	15	7.94	4.16
Social support (index)	1	4	3.28	.78	.95

Note: NSAHS, 2011–2014 = Nashville Stress and Health Study (n = 1,252)

Table 2 reports unstandardized coefficients from robust linear regression models of self-esteem. As shown in Model 1 of Table 2, neighborhood disorder is negatively associated with self-esteem (p < .05). In more detail, a one-unit increase in neighborhood disorder is associated with a .112-unit decrease in self-esteem, conditional on the other variables in the model. Thus, we find support for our first hypothesis.

Table 2.

Robust Linear Regression of Self-Esteem (n= 1,252), Nashville Stress and Health Study (2011–2014).

	(1)	(2)	(3)	(4)
Focal Variables
Neighborhood Disorder	−.112 (.044)*	−.110 (.044)	−.104 (.043)*	−.091 (.040)*
Organizational religiosity	−.025 (.017)	−.024 (.018)	−.024 (.018)	−.023 (.017)
Non-organizational religiosity	−.011 (.020)	−.011 (.020)	−.007 (.020)	−.013 (.020)
Perceived divine control	.109 (.040)**	.110 (.040)**	.111 (.039)**	.134 (.042)**
Interactions [Disorder x ...]
Organizational		.011 (.027)
Non-organizational			.056 (.025)*
Perceived Divine control				.198 (.077)*
Stressors
Recent life events	−.009 (.011)	−.009 (.011)	−.009 (.011)	−.011 (.010)
Financial strain	−.187 (.050)***	−.187 (.049)***	−.186 (.049)***	−.186 (.047)***
Socio-demographics
Age	.002 (.002)	.002 (.002)	.002 (.002)	.001 (.002)
Female	−.057 (.042)	−.058 (.042)	−.056 (.041)	−.051 (.040)
Black	.199 (.033)***	.198 (.033)***	.195 (.032)***	.186 (.031)***
Education	.003 (.007)	.003 (.007)	.002 (.007)	.002 (.007)
Married	−.001 (.039)	−.001 (.039)	−.001 (.039)	−.005 (.038)
Employed fulltime	.141 (.044)**	.141 (.044)**	.139 (.044)**	.137 (.043)**
Household income	.012 (.006)*	.012 (.006)*	.012 (.006)*	.013 (.006)*
Social support	.234 (.039)***	.234 (.028)***	.235 (.025)***	.232 (.038)***

Note: Unstandardized coefficients are reported with robust standard errors in parentheses. All estimates adjust for probability weighting and clustering at the census block group level.

^*p < .05

^**p < .01

^***p < .001 (two-tailed).

To test our remaining hypotheses, models 2 through 4 introduce interaction terms between neighborhood disorder and dimensions of religious involvement. We do not find support for Hypothesis 2, as organizational religiosity does not attenuate the inverse association between neighborhood disorder and self-esteem. However, we do find support for Hypotheses 3 and 4. The positive interactions in models 3 and 4 indicate that the negative association between neighborhood disorder and self-esteem is attenuated for respondents who score higher on private religious involvement and divine control, respectively. These patterns can be visually confirmed in Figures 2 and 3. These figures show that the inverse association between perceived neighborhood disorder and self-esteem is larger for respondents who score lower on private religiosity and divine control.

Table 3 reports average marginal predictions confirming that neighborhood disorder is significantly and negatively associated with self-esteem at all levels of perceived divine control and private religiosity. However, the slope coefficients gradually taper in magnitude as levels of religiosity increase.

Table 3.

Average marginal predictions, Nashville Stress and Health Study (2011–2014).

	dy/dx	s.e.	p
Non-organizational Religiosity:
−1SD	−0.415	(.161)	*
Mean	−0.342	(.129)	**
+1SD	−0.268	(.098)	**
Perceived Divine Control:
−1SD	−0.847	(.308)	**
Mean	−0.697	(.250)	**
+1SD	−0.547	(.193)	**

^*p < .05

^**p < .01 (two-tailed).

Ancillary Results

Because we used composite measures of organizational and non-organizational religiosity, some readers may be curious about whether the patterns remained the same when we substituted individual components of these indices when estimating our stress-buffering models. In ancillary analyses, we tested interactions between perceived neighborhood disorder and the separate measures comprising our public and private religiosity scales. Results of all ancillary models were substantively similar to our main analyses. In addition, to insure that the null findings involving the possible stress-buffering effects of organizational religiosity were not affected by the inclusion of general social support in our analyses, our models were re-estimated with this variable dropped. Key findings were unchanged by this re-specification.

DISCUSSION

A substantial body of research has linked neighborhood characteristics with mental health outcomes and psychosocial functioning. Our study has focused on perceived neighborhood disorder, which refers to conditions in which routine mechanisms of social control are understood to have broken down. The degree of neighborhood disorder can be gauged both by physical indicators (e.g., dilapidated buildings, overgrown lots, dirt and noise, etc.) and by social ones (e.g., the presence of petty crime and drugs, loitering, etc.) (Ross and Mirowsky 1999). Perceived neighborhood disorder focuses on the lived experiences of residents, i.e., their definitions of what is in/outside their neighborhoods (as opposed to Census tract data or some other arbitrarily defined areal unit), and the extent to which they perceive that there is a breakdown of conventional mechanisms of social control and regulation. Several specific findings deserve discussion.

Our first main finding is that neighborhood disorder is inversely associated with self-esteem, a conclusion that resonates with several earlier studies (Haney 2007; Hill et al. 2013). At the outset of this paper, we identified several possible reasons for such an empirical pattern. First, to the extent that perceptions of global moral self-worth –the essence of self-esteem—are influenced by reflected appraisals, as classic work by Rosenberg (1981) suggested, individuals residing in disordered neighborhoods may be treated negatively by neighbors, other associates, prospective employers, police and public officials, and others based on where they live. Thus, as others treat them as unworthy and inferior, they may internalize this negative self-image. Second, to the extent that residents recognize –through personal experience, media exposure, or other means—that their neighborhoods are characterized by visual and social cues of disorder, and are more troubled than other residential contexts in their wider community, they may find it increasingly difficult to maintain a positive self-image. Third, to the extent that self-esteem is shaped by perceived efficacy, residents of disordered areas may feel unworthy because they are unable to improve conditions in their neighborhoods and may find it challenging to secure alternative, more satisfying places in which to live.

Although the replication of such findings is useful, a more important task here is to assess whether multiple dimensions of religiosity moderate the links between neighborhood disorder and self-esteem. Several findings warrant brief discussion. First, our work builds on a number of recent studies that highlight the role of religious cognitions in shaping mental health and positive psychosocial functioning (Bradshaw and Kent 2018; Ellison et al. 2014; Schieman et al. 2017; Silton et al. 2014). After years of purposeful neglect, scholars from multiple disciplines are taking religious beliefs and experiences seriously, as potentially independent influences on mental health outcomes. A small but growing body of work centers on the sense of divine control, or the perception that God is playing an active role in directing the course of one’s life and managing one’s affairs (DeAngelis 2018; DeAngelis and Ellison 2017, 2018; Schieman et al. 2005, 2006, 2010). Indeed, several studies now report that such divine control beliefs moderate the links between chronic stressors and aspects of well-being (e.g., DeAngelis and Ellison 2017; Upenieks et al. 2022).

Why might the sense of divine control mitigate the link between neighborhood disorder and self-esteem? One set of reasons stems from that broader literature on religion and well-being. Viewed broadly, believers in divine control are better able to reframe stressors like neighborhood disorder in terms that are less threatening to their core sense of self (DeAngelis and Ellison 2017). For example, when seen through such a religious lens, unpleasant conditions such as neighborhood disorder can be interpreted as part of God’s plan, as opportunities for spiritual growth, or as relatively minor challenges that are small indeed when viewed alongside divine promises of eternal salvation (Pargament 1997). To those who believe that God is all-powerful and has a plan for their lives, the challenges and indignities posed by breakdowns of social control within neighborhoods –which constitute the essence of neighborhood disorder—are likely to seem more manageable and less threatening to one’s intrinsic self-worth. Thus, our findings regarding the stress-moderating implications of divine control belief fit well within the wider conceptual discussions of: (a) secondary control strategies, in which individuals may modify their self-perceptions to deal with conditions that lie beyond their control; (b) interpretive control strategies, which involve developing benign rationalizations of problematic circumstances; and (c) vicarious control strategies, in which persons may align themselves with potent exogenous actors to augment their own sense of self-esteem (Rothbaum et al. 1982).

A second possible explanation for the stress-buffering pattern involving the sense of divine control may reflect beliefs about God’s active role in protecting the faithful from harm. After all, persons with strong divine control beliefs endorse a view of God who directs the course of their own lives and helps to manage their affairs. This deity is typically loving and benevolent in nature, and the faithful may be inclined to trust God to shield them and their loved ones from physical or emotional harm stemming from the conditions in their neighborhoods. Although this issue cannot be explored directly with NSAHS data, a good example of how this sense of divine control might work can be drawn from recent literature on religious beliefs and COVID-19 vaccine uptake. Studies in this vein have shown that persons who believe in an active, beneficent God may eschew vaccines as unnecessary, since God’s willingness to intervene suffices to protect the faithful (e.g., DiGregorio et al. 2022). Using a similar logic, believers in divine control who reside in disordered neighborhoods may feel protected against discrimination, slights and insults based on place of residence, criminal victimization, and other hazards of daily life in such settings.

Yet a third possible explanation for the observed stress-buffering role of divine control belief in our study may stem from orthodox Christian beliefs, which are common in places like Davidson County, TN, where the NSAHS sample was drawn. Briefly, many religious persons believe in a benevolent and loving God who (a) created them in His image, (b) seeks a unique personal relationship with each member of His creation, and (c) offers eternal life to those who accept His offer of divine grace and forgiveness. This belief that one is special and significant in the eyes of God, and indeed, is loved and valued by God, may serve as a potent source of self-worth (DeAngelis 2018; Ellison 1993; Krause 2009; Schieman et al. 2017), and this may matter even more among persons who are coping with the psychosocial effects of chronic strains such as neighborhood disorder (DeAngelis 2018; DeAngelis and Ellison 2018). In addition, there is at least some evidence linking belief in divine control with the sense of mattering, or “inferred significance.” Thus, persons who perceive that God has authority over the course and direction of their lives tend to recognize their contributions to the world and to the lives of others, and they tend to believe that others would miss them in their absence (Schieman et al. 2010). Among persons living in disordered neighborhoods, such perceptions could be especially important sources of positive reflected appraisals.

In addition to the sense of divine control, our results also underscore the importance of non-organizational forms of religious practice in moderating the association between neighborhood disorder and self-esteem. Specifically, we gauged non-organizational religiosity in terms of (a) the frequency of prayer and (b) the frequency with which individuals draw upon their faith to deal with problems. On average, individuals who engaged in these non-organizational practices more often experienced fewer deleterious effects of neighborhood disorder than persons who prayed or turned to faith when coping less often. These patterns are consistent with a growing literature demonstrating the importance of religion –particularly prayer—in coping with personal challenges (e.g., Krause 1998; Pargament 1997; Pargament et al. 2000). Such research often shows that individuals derive psychosocial benefits from constructing and maintaining intimate personal connections with God, especially when they are confronting chronic or acute stressors (Krause 1998; Pargament 1997; Sharp 2010). Our findings add to this body of evidence, showing that prayer and coping buffer the deleterious association between neighborhood disorder and self-esteem.

We observed these significant patterns involving non-organizational religiosity despite having to rely on rather rudimentary measures of this domain. The NSAHS instrument inquired only about how often individuals pray or turn to faith when coping with problems. Researchers have documented a rich array of prayer styles and modalities (Froese and Uecker 2022; Krause and Chatters 2005; Poloma and Gallup 1991). Findings to date seem to show that regular prayer is especially beneficial for psychological well-being in the context of (a) benevolent God images, (b) secure attachment to God, and (c) certain types of prayer expectancies (Ellison et al. 2014; Upenieks 2022). Although we cannot adjudicate these issues with NSAHS data, it is certainly conceivable that the buffering effects of prayer on the links between neighborhood disorder and self-esteem could be even stronger under the conditions specified above. In addition, the seminal work of Pargament and colleagues (Pargament 1997; Pargament et al. 2000) has discovered a diverse array of religious coping styles and practices, some of which are much more beneficial than others. In particular, persons who experience God as a “force multiplier,” who augments their own problem-solving capabilities, fare particularly well according to their studies. Future research including more prayer dimensions (Froese and Uecker 2022; Upenieks 2022) and more religious coping measures (Pargament et al. 2000) could clarify and extend the findings reported here.

In addition to this evidence regarding apparent stress-buffering effects, yet a third pattern –this one involving a null finding—is also noteworthy. Briefly, we observed no significant role of organizational religious involvement –measured in terms of religious attendance, socializing with coreligionists, and support from church members—in moderating the association between neighborhood disorder and self-esteem. This null result is somewhat surprising at first glance, given the number of studies in the literature extolling the psychosocial benefits of congregational integration and support, both formal and informal. According to some studies, these facets of organizational religiosity are especially potent for persons who are dealing with stressful events and conditions, such as social or economic marginality and related strains (Krause 2003, 2008; Krause and Hayward 2014).

We see several plausible explanations for these counterintuitive null findings. First, it is conceivable that organizational religiosity –as opposed to non-organizational practices and beliefs (e.g., Krause 1998)-- offer no particular benefits to persons living in disordered neighborhoods, at least no more so than to other adherents. One potential reason for this involves higher rates of property crime and petty offenses (e.g., drugs, graffiti, etc.) that may afflict congregations located in troubled neighborhoods, constraining their reach into the lives of the surrounding community and church membership. Recent research has shed valuable new light on the extent to which religious congregations in certain types of areas –likely including those characterized by high levels of disorder—are victimized by crime, and the extent to which they may be forced to reduce access and compromise routine activities to limit their exposure (Corcoran, Scheitle, and Hudnall 2020; Ulmer and Scheitle 2020). Such criminal victimization and its consequences may deter some members from participating in congregational activities regularly, and it may hinder the churches’ ability to offer opportunities for social engagement and effective support systems for members. Consistent with this logic, ancillary analyses of the NSAHS data (not shown, but available upon request) revealed a modest inverse correlation between neighborhood disorder and frequency of church attendance (r=−.065, p<.001). Although this line of argument offers a speculative explanation for the patterns observed in our data, we believe it is promising and deserves further investigation.

Second, as noted above, at least one recent study highlights the role of individuals’ attachment to God in shaping their congregational engagement. According to that investigation, persons with a secure attachment to God tend to maintain congregational involvement and church-based social ties over time, whereas those with anxious or avoidant attachments to God reduce their social and religious engagement with their congregation and fellow church members (Kent and Henderson 2017). Such contingent patterns could well complicate the link between organizational religiosity and outcomes such as self-esteem. Although this issue lies beyond the scope of the current paper, it bears exploration in the future.

Yet a third explanation for the null evidence of any stress-buffering effects of organizational religiosity involves possible shortcomings in the measurement of this construct in the NSAHS data. The NSAHS includes only one item on church-based social support, asking respondents “how often” members “help [them] out...“ More extensive, nuanced measures of congregational integration and support have been incorporated by other researchers, who routinely distinguish between (a) socioemotional vs. instrumental or tangible support; (b) support from church members vs. support from clergy; and (c) receipt vs. provision of support (Chatters et al. 2018; Krause 2008; Nguyen et al. 2017). Although we found no stress-buffering effects using our single, omnibus congregational support item, it is possible that use of a more detailed battery of items on church-based friendships and support exchanges could yield different results (but see Krause 2009).

Like most studies, our work is characterized by several limitations. First, our data are cross-sectional, and therefore temporal order among focal variables cannot be established. This makes it impossible to assert strong causal claims. For example, some early discussions of the links between neighborhood conditions and psychosocial functioning also suggested that reverse causation could be at work, muddying this association. In other words, persons with low self-esteem or other psychosocial deficits could be particularly prone to select into –or might be channeled into—troubled neighborhoods due to a myriad of unmeasured factors. Our data do not permit us to adjudicate such issues. In addition, the data come from a sample of African Americans and non-Hispanic Whites residing in one metropolitan county in the urban south (Davidson County, TN). Respondents range from 22 to 69, thus excluding older adults. Generalizations to other racial and ethnic subgroups, ages, and types of communities would be unwise. Further, as discussed above, although our measures of core religious domains are adequate, richer and more variegated measurement approaches could conceivably yield different results and divergent conclusions. In addition, although there is little, if any, evidence that self-esteem varies systematically by religious affiliation, it would have been optimal to have a denomination/affiliation measure in the NSAHS. Future studies should explore this issue if possible. In addition, the stress-moderator patterns observed here may, in turn, be contingent on one’s place in the stratification system. Additional work to explore race- and education-based contingencies could be fruitful (DeAngelis and Ellison 2018; Krause 2008, 2021; Upenieks and Schieman 2021). Further, this study has focused specifically on neighborhood disorder, gauged via subjective perceptions of one’s neighborhood and signs of the breakdown of social control and order within it. Although this is a valuable way to investigate neighborhood conditions, many social scientists have employed other means, including the use of information on Census tracts, or administrative units designed to be relatively homogeneous in population size and composition, and are intended for official data collection and analysis. In the future it would be useful to explore the associations between Census tract characteristics (e.g., poverty and crime rates) and individual self-esteem, and the possible role of religious factors in mitigating the effects of neighborhood disadvantage measured in these and other ways.

Moreover, our study has focused exclusively on potentially salutary (i.e., stress-buffering) roles of multiple religious domains. Religion may be implicated in the association between perceived neighborhood disorder and self-esteem in other ways as well. For example, some religious and spiritual domains –e.g., spiritual struggles—have been shown to be associated with undesirable psychosocial outcomes. Three aspects of spiritual struggles have received much of the coverage in recent studies: (a) divine struggles, or troubled relationships with God; (b) intrapsychic struggles, or chronic, unresolved religious doubting; and (c) interpersonal struggles, or conflicts that involve religious settings, institutions, persons, or topics (Exline et al. 2014). It is possible that perceived neighborhood disorder –and its visual and social cues—could contribute to spiritual struggles, which in turn to erode self-esteem. For example, the experience of neighborhood disorder could lead individuals to grapple in new ways with the idea that God has visited such chronic strains upon them, which could lead them to feel less worthwhile and competent as individuals (Hill et al. 2023). As an alternative to this potential mediating pattern, it is possible that spiritual struggles could exacerbate the already problematic link between neighborhood disorder and self-esteem. Such a pattern has been observed in studies of other stressful events and conditions (e.g., chronic health problems) and other psychosocial outcomes (e.g., psychiatric symptoms) (e.g., McConnell et al. 2006). These alternative possibilities and models merit attention from investigators in the future.

Such limitations notwithstanding, our study casts fresh light on the linkage between perceived neighborhood disorder and self-esteem, and the moderating role of multiple religious domains. Our findings underscore the importance of non-organizational practices and religious cognitions –specifically, beliefs about the authority and presence of God in daily life—in buffering the otherwise deleterious associations between perceived neighborhood disorder and self-esteem. A number of researchers over the years have built on classical social psychological theories in explaining patterns and possible consequences of prayer and other forms of spiritual practice (Luhrmann 2012; Pollner 1989; Sharp 2010). At the same time, an emerging literature has focused on religious experiences and cognitions and their potential implications for psychosocial functioning, especially in the face of stressful events and conditions (Flannelly 2017; Park 2017). Although religious congregations and organizational religiosity will retain a central place in this line of inquiry, as social scientists continue to take non-organizational practices and religious cognitions seriously, these efforts will further illuminate the rich and variegated connections between religion and various facets of psychological well-being.