Loss of Trust in the Neighborhood: The Experience of Older African Americans in Detroit

Abstract

Objective

Although evidence suggests that neighborhood conditions are related to stress and health, the processes connecting neighborhood conditions and stress for older minorities is little explored. The purpose of this analysis is to contribute new insights into this issue.

Method

We conducted a qualitative analysis as part of a larger mixed methods study of 100 African Americans aged 55 years and older living in neighborhoods of varying quality in Detroit, Michigan. A subsample of (n = 20) older adults took photographs of bothersome aspects of their neighborhoods and participated in in-depth photo-elicitation interviews. Data were analyzed using a grounded theory approach.

Results

“Loss of trust in the neighborhood” emerged as the core category to explain how older African Americans in our sample experienced neighborhood stressors in their daily lives. Loss of trust in physical, social, and institutional dimensions of the neighborhood contributed to the core category.

Discussion:

The life course of neighborhoods and the trust placed in them appears to be intimately connected to the well-being of older African Americans. We therefore hypothesize that a fundamental pathway through which neighborhood stressors are experienced for older African Americans in United States “Rust Belt” cities is the multifaceted loss of trust in the neighborhood.

Background

The growth of social epidemiology during the last several decades was spurred on by stress research, a return to ecological thinking in epidemiology, and the rise of population health and a related health disparities focus during the 1990s (Berkman & Kawachi, 2000). An important unit of analysis in the field is the neighborhood (Kawachi & Berkman, 2003), and research on neighborhoods and health has grown exponentially during the last 20 years (Diez Roux & Mair, 2010). Stress remains a frequent target of inquiry because of its position between health outcomes and neighborhood stressors, and this has become a more significant focus of gerontological inquiry as well (e.g., Everson-Rose et al., 2011; Geronimus et al., 2015; Lehning, Smith, & Dunkle, 2014).

Urban neighborhoods are an important unit of analysis when examining health and place relationships because of the variation in, and clustering of, a range of stressors (e.g., poverty, disorder, crime). Various factors and dynamics at the neighborhood level are posited to act as stressors. Social, cultural, and psychological processes such as unfair treatment (Schulz et al., 2012), social disorder (Ross & Mirowsky, 2001), and neighborhood satisfaction (Coulon, Wilson, Alia, & Van Horn, 2016) are believed to act as stressors or mediators. Although studies have identified a range of physical and social neighborhood features associated with stress, health behavior, and health outcomes, the complexity of the subject matter and the need for better methodological approaches to improve inferences are ongoing concerns (Clarke et al., 2014; Diez Roux & Mair, 2010).

Neighborhoods also are a point of intersection for larger societal influences, local regulations, social practices, and individual behaviors. Because policies, practices, and individuals change through time there is dynamism of neighborhood life that is contingent on history and experience in place (Cutchin, 2007; Keene, Bader, & Ailshire, 2013), and that varies across race and ethnicity (Schulz et al., 2008), relative social position (Kelley-Moore et al., 2015), and gender (Bassett & Moore, 2013). We argue that this dynamism is important for understanding the neighborhood circumstances that predispose some individuals or groups to experience stress while others exhibit resilience. Qualitative methodologies are uniquely suited to access the complexity of how people experience their neighborhood and they can help develop needed theory about neighborhood and health processes (Diez Roux & Mair, 2010; Gardner, 2011). Scholars have previously suggested that using qualitative methods to produce hypotheses, data, and inferences driven by the experience of residents themselves could enhance our efforts to identify ways to improve neighborhood environments (Israel et al., 2006). With few exceptions (e.g., see Al-Bayan, Islam, Edwards, & Duncan, 2016; Fritz & Cutchin, 2017; Turney, Kissane, & Edin, 2013), the literature contains very little knowledge based on the experience of residents and their views relevant to their neighborhood conditions and stress, and knowledge is even more limited regarding older African American residents.

Graham Rowles was a gerontological pioneer in attending to the importance of neighborhoods and the way they are experienced by older people as important spaces and places in everyday life. His qualitative research in an inner-city working-class neighborhood of the Northeast United States discovered how older residents accommodated neighborhood change through various adjustments to increasingly distressing spaces and places in the neighborhood, such as changing patterns of action, feeling, and fantasy (Rowles, 1978, 1980). Gardner (2011) provided a cogent argument for the value of attending to neighborhoods in studying older adult well-being:

Neighborhoods are an important and meaningful context of life for older people aging in place. With fewer responsibilities requiring them to go beyond their proximal spaces, and with functional limitations that can make it difficult to do so, many older adults spend increasing amounts of time immersed in neighborhoods where they have often lived for an extensive period of time (p. 264).

The connections between neighborhoods, aging, ethnicity, and stress is additionally important. While there are various theoretical positions on stress relevant to draw from to assess such connections, the stress process model is particularly well situated. A primary objective of that model is to explain how low social status and relative lack of resources leads to stress and poor mental health (Aneshensel, 2009; Pearlin, 1989, 1999). Neighborhoods in U.S. cities are an especially important social context in which to examine the stress process (Aneshensel, 2009). In the United States, a legacy of structural racism, segregation, and economic disinvestment has resulted in older minorities disproportionately inhabiting, and therefore aging, in stressor-laden urban neighborhoods (Massey & Denton, 1998; Massey, 2001). There is evidence that older adults may engage with and experience neighborhood conditions in different ways than their younger counterparts (Becker, Israel, Schulz, Parker, & Klem, 2005; Tuttle, Meng, Moya, & Johns, 2012; Yen, Michael, & Perdue, 2009). At the same time, because of their potentially longer-term relationship with their neighborhoods, changes in multiple physiological systems, and reductions in social supports, the synergistic effects of aging in disadvantaged neighborhoods may be more problematic for the health of older residents (Alishire, Karraker, & Clarke, 2017). For these reasons, eliciting the experiences of older minorities may be especially significant for understanding how history and experience in place contribute to neighborhood and health relationships, especially in the context of aging.

The purpose of this article is to contribute new insights into how older minorities experience their neighborhoods as potentially stressful places and to suggest processes linking neighborhoods and health. We present a qualitative analysis from a study examining the relationship of neighborhoods and stress for older African Americans living in Detroit, MI. Neighborhood change is a constant everywhere, but the socioeconomic context of neighborhood change in Detroit is unique in its intensity. At the same time, Detroit shares the roots of change with other “Rust Belt” cities in the United States whose factories and associated jobs faded away during the latter part of the 20th century. Detroit was historically racially and ethnically diverse because of labor migration for employment in the auto industries (including the “great migration” of African Americans from the south), but it also was very segregated and plagued by racialized politics (Sugrue, 2006; Thompson, 1999). A large African American middle class emerged by the mid-20th century, but as auto industry jobs declined and “white flight” to the suburbs accelerated in the 1970s, Detroit became a majority African American city (83% in 2016) with only one-third of the total population of the early 1950s (Sugrue, 2006; Thompson, 1999; worldpopulationreview.com). Therefore, while Detroit had been in economic and population decline for several decades prior to the “Great Recession”, many of those long-term problems and socioeconomic distress became intensified during the financial and housing crisis of 2008. As such, change in Detroit neighborhoods accelerated, often for the worse. The culmination of this latter phase of change was the city’s bankruptcy in 2013 (Kruman, 2017). We hypothesized that because of the timing and context of neighborhood change in Detroit, residents would be primed to discuss their neighborhoods and associated stress as well as provide important insights to help address the gaps in knowledge about older minorities.

Methods

The analysis presented here was one component of a larger mixed-methods study examining the relationship among daily activities, neighborhood environmental stressors, and the body’s stress response (measured via salivary and hair cortisol) among 100 African Americans aged 55 years and older living across a wide range of neighborhoods in Detroit, MI.

Sample and Recruitment

All data were collected between September 2015 and September 2016. To be included in the study, participants had to self-identify as an African American aged 55 years or older living within the city of Detroit and be able to participate in daily activities in the neighborhood environment without physical assistance from another person. Our target sample of 100 older African Americans was initially recruited by way of stratified random sampling from the Wayne State University Institute of Gerontology’s Participant Resource Pool (PRP). The PRP is a volunteer registry of 1,400+ African American individuals aged 55 years and older willing to participate in research of interest to them (Chadiha et al., 2011). Five strata of neighborhood adversity were created by a neighborhood (census tract) disadvantage index. That index was composed of the sum of three measures collected from the 2010 U.S. Census and the 2009 Data Driven Detroit Residential Parcel Survey: proportion of housing units in less than fair condition, poverty rate, and proportion of unoccupied housing units. We augmented random sampling with snowball sampling (participant referral of other potential participants). The latter approach contributed approximately a third of the full study sample. The Wayne State University institutional review board approved the study protocol.

Our goal was to have 20% of the larger sample of 100 older African Americans participate in the qualitative component (N = 20) and to maximize variation across the sample based on our five strata of neighborhood adversity within the qualitative sample. Specifically, during the informed consent process for the larger study, participants were asked if they would be willing to participate in the additional qualitative component (if so, informed consent was obtained for the additional qualitative component). We invited participants into the qualitative component in the order that they were enrolled into the larger study until we met our recruitment goal of 20 participants for the qualitative component. We made efforts to recruit qualitative participants from each of our five neighborhood strata to maximize neighborhood variation as well as gender variation.

Sample Characteristics

Despite efforts to recruit equal numbers of males and females into our qualitative sample, fewer males agreed to participate both in the larger study and in the qualitative component (see Table 1). Participants were all self-described life-long Detroiters, with the majority (n = 12) living 20 or more years in their current neighborhood. The average age of the qualitative sample was 72 years. Five participants lived in the least adverse (best) neighborhood stratum, 10 lived in the second best, and 5 lived in the remaining lower neighborhood strata. Nineteen different census tracts (neighborhoods) were represented. About half of the sample lived alone while the remainder lived with at least one other person. Fourteen participants had some college education or technical school and of the 17 participants who provided data about household income, 13 reported an income between 10K-39K per year.

Table 1.

Sample Characteristics

N = 20	Frequency or M (SD)
Gender
Female	16
Male	4
Age	72 (7.4)
Census Tract Strata
Best	4
Second best	7
Middle	4
Second worst	3
Worst	2
Relationship Status
Never married	3
Married	5
Separated or divorced	9
Widowed	3
Household Size
1	11
2	4
>2	5
Education
High school degree or equivalent	3
Some college or technical school	14
College graduate	3
Annual Income
Less than $10,000	1
$10,000-$19,999	5
$20,000-$29,999	3
$30,000-$39,999	5
$40,000-$49,999	2
$70,000+	1

Data Collection: Participant Photography and Photo-Elicitation Interview

Data for the qualitative component of the study were collected through participant generated photographs and photo-elicitation interviews. Photo elicitation interviews have been used widely in social science research and are an effective way to collect data on contexts, subjective experiences, and tacit processes across a variety of settings, populations, and problems (e.g., Collier, 1957; Harper, 2002). Photo elicitation interviews differ from other visual methodologies such as photovoice. For example, photovoice is a more prescriptive photo methodology framework used within community based participatory action research projects explicitly for the purposes of fostering community participation, involving local stakeholders, and reaching policy makers (Wang & Burris, 1997). The photo elicitation approach, in contrast, is a flexible method that, in its most basic form, involves the use of photographs in conjunction with research interviews (Harper, 2002). Photographs serve as the catalyst for discussion, but the resultant narrative may or may not map directly onto the visual content of the photographs.

Participants were already using an Android smartphone for seven consecutive days to collect activity data as part of an ecological momentary assessment component of the larger study (Fritz, Tarraf, Saleh, & Cutchin, 2017). They were instructed on the use of the smartphone’s camera function and asked to take photographs during the same 7-day period. Participants were given the following prompt: “Take photographs in your community while you are out doing your usual everyday activities. The photographs can be of anything in your neighborhood that you feel is important to your daily activities or stress. Some ways to think about stress is to think about things in your community that you feel are irritating, frustrating, make you feel unsafe, or are just things that bother you.” We did not attempt to define “neighborhood” for our participants. Instead, participants were free to interpret the concept of neighborhood on their own terms. Because participants were instructed to not take photographs of their own or others’ faces, photographs were of places, spaces or buildings in the community that represented issues, people, or events (e.g., a photograph of a home with the curtains drawn that served as a catalyst for the participant to discuss her lack of trust for some of her newer neighbors).

After the 7-day data collection period, a research assistant returned to the participant’s home to collect the phone and scheduled a return visit for the photo-elicitation interview. Photo-elicitation interviews were scheduled within 14 days of the photography data collection period. Participants who took photographs and participated in the photo elicitation interview received $50 compensation for their participation. A total of 720 photographs were generated across the sample (range: 3–129; average: 36). For each participant, photographs were copied to an iPad or laptop to view during photo-elicitation interviews. The first author, who had previous experience performing home-based qualitative data collection and photo elicitation interviews, conducted all interviews. All interviews were conducted in the participant’s home. During the interview, the researcher and participant reviewed each photo taken by the participant.

Participants were asked to explain the content and meaning of each photograph (with the exception of multiple photographs taken of the same phenomenon) and to expand upon issues that were raised through those explanations. For example, one participant presented a photograph of an abandoned building and discussed a violent crime that had happen in the building 2 years earlier. In the process of explaining the photo, the participant commented on police responsiveness and was therefore asked to expand upon the comment in more depth. In addition, issues that arose from participants in early interviews (e.g., squatters) were queried in subsequent interviews if not brought up by the participants themselves. All interviews were audiotaped and transcribed. The average interview length was 50 min with a range from 23 to 120 min and resulted in 566 total transcribed pages used for analysis.

Data Analysis

Data were analyzed using a grounded theory approach (Charmaz, 2006). All three authors (a female occupational therapy researcher and gerontologist, a male geographer and social gerontologist, and a female sociologist), contributed to the analysis with the first and corresponding authors taking a primary role by conducting 80% of the open coding. All three authors are trained qualitative researchers with previous experience with qualitative data collection and conducting qualitative analysis; the first and corresponding author are trained and practiced in grounded theory analysis. Data analysis began with the authors open coding (coding small segments of text line-by-line) the interview transcripts. Following the open coding step, data were focus-coded which involved taking initial analytical directions developed from the line-by-line coding and applying them more selectively to larger chunks of text (Charmaz, 2006). The coding resulted in the extraction of 740 fragments of text and with between one and three applicable codes being assigned to each.

Memos were written and discussed by the team during the process of developing categories for the codes. The iterative process of category refinement and modification occurred until the categories seemed to best fit the data. To ensure trustworthiness of the analysis, the consolidated criteria for reporting qualitative research (Tong, Sainsbury, & Craig, 2007) were used when planning and executing this study. All data were coded and reviewed by two different coders. When differing viewpoints about the data arose, they were discussed and resolved by the team through consensus. Data and preliminary interpretations were not presented to participants during the analysis. Eventually, a core category was arrived at which seemed central to understanding the process by which older African Americans in Detroit experience neighborhood change and stressors in their daily neighborhood life. In the following section, we describe the core category and associated model, and we include participants’ quotes to illustrate main points (all names used are pseudonyms).

Findings

Loss of Trust in the Neighborhood

Loss of trust in the neighborhood is the core category that helps to explain how older African Americans experienced neighborhoods across Detroit in 2015. The core category refers to the decline in trust granted to individuals or groups in the neighborhood context, but the data also reveal that trust in the physical environment and local institutions is diminished and related to residents’ stress. Our data and findings suggest that neighborhood change experienced by our participants over a number of years is an important basis for the loss of trust and associated concerns, frustrations, and stress that seem associated with it. The core category of loss of trust has three dimensions that reflect the different neighborhood changes that contribute to a general loss of trust (see Figure 1).

Figure 1.

Model of loss of trust in the neighborhood.

Loss of Trust in the Neighborhood Physical Environment

While trust is most commonly associated with social relations and the amount of trust granted to individuals, groups, or organizations, we discovered the physical environment to be a significant issue in our participants’ sense of neighborhood trustworthiness. Participants discussed changes in neighborhood conditions as a significant concern and stressor. Unlike more stable conditions experienced in the past, changes to environmental conditions in their neighborhoods had created situations that were perceived as untrustworthy in multiple ways.

Photos and narratives included numerous examples of abandoned and decaying housing stock that, as suggested by Grace (age 84), were perceived as untrustworthy because they were associated with threats to self and property: “They [vacant properties] are not safe. Either there isn’t anyone living there or there’s squatters there, or they’re burned out, or gangs are hanging out in them.” Darcia (age 70) echoed Grace’s concerns and added that abandoned properties were especially unsafe for young people in the neighborhood stating, “I really don’t think it’s safe for the children when they’re going to school. There could be drugs in one of those houses or [they could be] raped, or raped and killed, the young ladies or boys too.” Certainly, the perceptions and loss of trust in neighborhood housing structures is informed by social perceptions, but the loss of trust also appears to be rooted in deteriorating infrastructure. As will be illustrated in subsequent sections, we hypothesize that much of the social and institutional dimensions of trust flow from the environmental changes to neighborhoods where our participants lived.

Equally concerning to residents in our sample were unmaintained lots and alleyways (see Figure 2 for an example), inadequate street lighting, and ad hoc dumping grounds. Participants associated overgrown lots and alleys with hazards such as wild dogs, rodents, and snakes. For example, when asked why the vacant lot near her home was a concern, Tamara (age 78) stated, “Because anything can run out of there, dogs, a man; it’s dangerous.” Terrance (age 66), while expressing concern about dogs, was also concerned about rodents and other vermin stating, “If you get the tall grass, you’ll have field mice and everything else. You have a chance they’ll come into your house.” Though most of the unkempt lots that residents encountered in their neighborhoods were privately-owned or bank-owned properties, the city was also reported as negligent in maintaining ‘greened’ lots and other green spaces such as parks. Althene (age 83) discussed a nearby city park, saying, “The grass is almost up to your knee sometimes before they [the city] decide to cut it, and you can’t enjoy the park because they won’t keep it maintained.” Poor street lighting compounded concerns about threats by further obscuring them during the nighttime. Many participants echoed the sentiment expressed by Dale (age 87) that poor lighting was “One of the most significant issues in the neighborhood.” The total darkness in some areas of the neighborhood made participants reticent to be outside after dark.

Figure 2.

Unmaintained lots and alleyways.

However, the loss of trust in neighborhood conditions went beyond safety concerns and pest control. Participants expressed that along with security and predictability, neighborhood conditions of the past had afforded them an expected level of comfort and aesthetic enjoyment that had been diminished so much that conditions now felt unsatisfactory, indeed untrustworthy, for many. The proliferation of ad hoc dumping grounds was especially stressful to some participants. Grace (age 84) provided a description of the situation in her neighborhood when discussing one of her photographs, “Look what is happening here. People are dumping. There is dumping everywhere. There is garbage, mattresses, and carpet. There’s all kinds of stuff right in the middle of the community!”

Through their past experience in their neighborhoods, participants in our sample appeared to have come to expect a certain level of trustworthiness in the physical environment. That trust had been eroded during the recent period of significant neighborhood change. In particular, the deterioration of their neighborhood’s physical environment led residents to lose trust in their environment in terms of providing reliable security, safety, and the previously expected aesthetic experience.

Loss of Trust in the Neighborhood Social Environment

Demographic change in neighborhoods also appeared to alter trust and induced stress for our older resident sample. The changes most important to participants were tied to evolving composition of neighborhood residents. Older residents sometimes “aged out” of the neighborhood because of illness, disability, or death. Subsequently, the trend was for younger residents to move into the neighborhood, sometimes as owners of properties but often as renters of properties that had been previously occupied by owners. The housing crisis of 2008 onward was a central dynamic of residential change that older residents viewed as the root of many problems in their neighborhoods.

Our older participants tended to express that their new neighbors—younger renters in particular—were particularly untrustworthy. Based on experience, and sometimes on the basis of stereotypes, our participants often associated younger renters with the deterioration of neighborhood quality. For example, Sherrie (age 72) associated her neighborhood’s deterioration with the influx of renters saying, “When they started renting the house out, that’s when the neighborhood started going [down].” Tamara (age 78) also perceived that renters contributed to neighborhood decline and expressed relief that her neighborhood included mostly homeowners when she stated, “We have been really blessed and lucky to not have those type of people in the neighborhood [low income renters]. ... I don’t know what happened in the next block, but in this block, they’re pretty well decent people.”

Our participants also shared more concrete evidence of concerns about their neighbors. A common cause of loss of trust in neighbors was based on a lack of adherence to perceived or expected social norms. Participants talked about how they could not trust new neighbors to uphold standards that they viewed as basic and reasonable. Janice (age 75) suggested that caring for one’s property was an important social norm that new neighbors had violated when she discussed a photo of nearby house (see Figure 3), “This house is now a rental and the porch over there, they constantly got some kind of heap of trash on that porch.”

Figure 3.

Rental house with trash on porch.

Grace (age 84) also expressed frustration with her new neighbors because she felt that they did not adhere to normal times and places for outdoor activities, “You don’t sit on your front porch and barbecue and have loud parties… Here you can’t blast music just like you’re trying to fill an auditorium for a thousand people to hear the music.” In addition to noise, participants also discussed that children were frequently outside without proper adult supervision. Additional problems were created when unsupervised children failed to uphold behavioral norms such as respecting other people’s property. One example of this was when Valerie (age 58) described an exchange that she had with some neighborhood children.

I was trying to grow grass and stuff, you know, and they were throwing a football. They were all over my grass. And it just pisses me off. I said “You all are going to have to get off my grass. Go over there and throw the ball. Why are you all over here?”

With the change in property ownership and associated change in the composition of neighbors, older adults appeared to lose much of the trust basis of neighborly relationships. Relationships with neighbors sometimes became strained, and new trust relationships became more difficult to develop. Trust had eroded so much for some that that interacting with new neighbors was considered a risk not worth taking; to notify or confront a neighbor about a problem might risk harm to self or property. Danny (age 82) summarized those concerns and explained why they led him to avoid engaging with his neighbors,

It’s very stressful. It was supposed to be a senior building, but they had all types of low income people living there, and different ages . . . Across the hall, somebody, they sell drugs so all the traffic that comes in and out is that. And we hear them banging on doors, banging on doors, all the time here, so…[Do you ever say anything to them?] No. I’ve thought about that. You couldn’t cause they would know where you live.

Intertwined with the shifts in the landscape and the associated housing crisis in Detroit’s neighborhoods was the erosion of social trust relationships based on the estimated trustworthiness of newer neighbors. In addition to a general bias against the change that younger neighbors appear to signify to older residents, the loss of previously experienced social norms—expectations of neighborly behavior based on previous experience in the neighborhood—was shared by our participants, regardless of neighborhood. The decline in trust of neighbors and a perceived inability to correct deteriorating social conditions also appeared to be significant detriment to participants’ quality of life.

Loss of Trust in Institutions

Neighborhood life entails more than the physical and social dimensions of the environment already described. Local and extra local institutions (e.g., government agencies and banks) have a large role in the development, upkeep, and repair of services and infrastructure in urban neighborhoods. Our participants explained how the City of Detroit sometimes responded to and solved problems, but the City also was viewed as an unreliable institution that often showed little interest or ability to help. For example, a very common problem was the need for functioning streetlights, many of which were not working and not being repaired. Neighborhood streets were often dark, and participants noted that they felt their neighborhoods were more dangerous as a result. In addition, participants discussed the crumbling sidewalks and roads and how the failure of the city to repair them placed older adults at increased risk for falls or accidents when going about their daily activities. Even when older adults contacted the city repeatedly about issues, the issues often remained unresolved. Lydia (age 87) expressed her frustration with the water department to resolve a leak at the vacant property next door to her home. She said, “They tell you to have everybody [in the neighborhood] call, blah, blah, blah. And then you might get some results in a year. You almost got to worry them to death.”

Nonpublic institutions also had become less trustworthy in the eyes of our participants. Some participants openly blamed nonpublic institutions for the downturn of the neighborhood, and those comments were most often directed at banks and mortgage lenders. Danny (age 82) provided an example of this when describing the banks role in his neighborhood’s decline. “I don’t think it’s by accident that we had these foreclosures, where people got these variable rate mortgages. I think when the bank gave them those mortgages, they knew those people couldn’t handle the accelerating interest rates.” Banks and real estate companies were viewed by some participants as guilty of letting properties decay, letting lots become overgrown, or generating the context for an influx of transitional residents. One example heard from several participants is that renting foreclosed properties without performing background checks resulted in bad neighbors moving into the neighborhood. Moreover, participants often expressed frustration at the limited recourse they and their neighbors had to address such issues. Neighborhoods are dependent on institutions to address many basic needs, and in Detroit during the past decade key institutions have been weakened and lost the trust of neighborhood residents.

Summary

Loss of trust in the neighborhood is a dynamic process that was common across our older African American participants in Detroit. The physical environmental, social environmental, and institutional dimensions of loss of trust have discernable roles in the larger core process. We note, however, that in the data we collected and analyzed, the dimensions often were codeterminant and co-constitutive and had direct and indirect influences on the core concept of loss of trust in the neighborhood; and those three dimensions were therefore central to our hypothesized model of how neighborhoods become stressful for older African Americans. The process of loss of trust in the neighborhood appeared to have taken place over the later part of the life course of our participants, and trust appeared to erode more after the financial and housing crises of 2008. Therefore, the backdrop of the loss of trust in the neighborhood, and a significant influence on the changes in physical, social, and institutional aspects of neighborhood environment, were the constellation of larger socioeconomic changes in Detroit, which in turn were linked to national and global disruptions in the economy. The model of loss of trust in the neighborhood for African Americans in Detroit, therefore, is partially contingent on a particular time and place—the situation of being an older, disadvantaged minority in neighborhoods within a city that has been relatively poor, beset with problems, and especially challenged to respond to such disruptions. Concomitant with that situation and contingency, however, is the fact that Detroit is not alone in such struggles or completely unique with its neighborhoods in transition inhabited primarily by older African Americans. The model does not intend to suggest that the processes described are uniform over time and place, nor does it suggest that trust cannot be redeveloped in response to such loss.

Discussion and Conclusion

The limitations of this study include its sample size and geographically limited urban context, and majority female sample. Research examining the relationship between neighborhood change and stress in other contexts or among a majority male group of participants may yield different inferences. Nonetheless, the sample was reflective of many different types of personal and neighborhood situations and the methodology allowed in-depth and multiple forms of data to be analyzed. The threads through those situations were consistent enough to infer the themes that defined the process of neighborhood change and loss of trust. We should note that data about the change of neighborhoods over time was gathered through participant narratives and not through objective means. We also acknowledge that the photographs used, while advantageous in allowing participant control of the subject matter and priming participants to think prior to the interviews, were generated over a 1-week period, which might limit the range of issues and dynamics elicited in the interviews. The photo prompt also suggested images about things in the neighborhood, generating evidence for the interviews that was focused more on material aspects of the neighborhood than might have been otherwise. Photos also cannot capture other stressors in neighborhood environments, such as noise and social interactions, which may have created further bias in our data. Such omissions might be captured in future studies by including daily diaries, for example, to expand the scope of understanding of neighborhood adversity and associated experience.

Our grounded theorization’s core category, loss of trust in the neighborhood, emerged from our data and analysis, but it is not entirely new to the literature on neighborhoods and health. A qualitative study by Walker and Hiller (2007) also found social trust relationships were important in how older Australian women perceived their neighborhoods and personal well-being. The seminal work of Ross and colleagues, working in the stress process tradition (Aneshensel, 2009), attended to the relationship among neighborhood disadvantage, disorder, and mistrust. Ross, Mirowsky, and Pribesh (2001) offer a theory of structural amplification of threat and found support for it in a large sample of Illinois residents, arguing that neighborhood disadvantage is the root of neighborhood disorder, which in turn causes social mistrust—the loss of neighborly relations as well as a sense of control and confidence to change the situation. Ross (2011) later extended the theory by suggesting that mistrust is an essential mediator of the neighborhood disorder and stress relationship. While Ross’ theory resonates closely with our findings, this investigation is among the first to use qualitative evidence from older adults about their neighborhoods and stress. In doing so, our data suggest additional insights about the loss of trust in neighborhoods.

The literature to date on neighborhood disorder and trust has focused almost solely on social dimensions of trust. In Detroit with older African American residents, we discovered that the physical/built environment and institutions are also important domains of trust and its loss through neighborhood disorder. Our domain of loss of trust in the physical environment is supported by previous qualitative and quantitative work linking physical neighborhood conditions such as abandoned buildings and decaying infrastructure to both perceptions of stress (Israel et al., 2006; Garvin, Branas, Keddem, Sellman, & Cannuscio, 2010) and health outcomes (Cohen et al., 2003; Schulz et al., 2013). The recent contribution by Garoon, Engelman, Gitlin, & Szanton (2016) suggested that deterioration of the neighborhood built environment was related to an instability of trust for older African Americans in Baltimore. Moreover, those authors argue that their participants’ lack of trust in the system and its historic bias against them was part of the mistrust in others. In our data, the lack of engagement of city institutions in the neighborhood since the financial crises was a much more pronounced aspect of loss of trust in our participants’ narratives than systematic racism or bias—although racism is likely an important part of our participants’ past and present experience, we did not focus upon it explicitly in our photo protocol or interviews and cannot draw detailed conclusions about racism and neighborhood stressors. Our participants’ photographs and narrative depictions of the declining trustworthiness of their neighborhood environments adds significant understanding of how disorder is manifested materially and produces stress. We note that similar to the recent findings of Portacolone, Perissinotto, Yeh, and Greysen (2017), the effects of a deteriorating physical environment can be stressful by diminishing social interaction and other forms of healthful activity (Fritz & Cutchin, 2017). While institutional and social trust issues are interwoven with the physical, the important role of the physical dimensions of trust and its loss adds important insights into neighborhood disorder, stress, and health—and it helps to overcome some of the limitations of social epidemiology by proposing a way to think about how neighborhood landscapes are therapeutic or not (Cutchin, 2007).

We also need to emphasize the temporal and comparative basis of trust valuations by our older participants. In agreement with Garoon et al.’s findings of trust being variable over time, our data indicate that the loss of trust in the neighborhood is dependent upon how residents view their physical neighborhood conditions and social connections to neighbors in relation to how they used to be. One important dimension of the then-versus-now comparison is the change in the perceived “type” of person now living in the neighborhood. Many participants viewed new residents as different from themselves (and former neighbors) because of their younger age, but also because of their perceived relative socioeconomic status. Participants used phrases such as “section eight” or “low income” to describe new neighbors, terms that they did not utilize in their descriptions of their previous neighbors. Some research (Bjornstrom, 2011) suggests that individuals with higher relative socioeconomic status are more likely to mistrust lower income neighbors. It is possible that participants’ mistrust of new neighbors was similarly shaped by their perceived higher relative socioeconomic status compared to that of their new neighbors.

Change in the neighborhood and how older residents view their neighborhood, therefore, becomes a fundamental dynamic in the maintenance or loss of trust. Living in a neighborhood that does not reflect the values and quality that it once had earlier in the life course, and having little control over how to modify the situation, may increase the stress-inducing potential of the physical and institutional dimensions of the loss of trust. Rather than call the core category “mistrust,” loss of trust better reflects this temporal dimension of neighborhood evolution and experience—especially in the case of our Detroit sample after the financial crisis of 2008. Through the grounded theorization, it becomes more apparent how the life course of neighborhoods and the trust placed in them appears to be intimately connected to the well-being of older African Americans.

Scholars have questioned the use of trust as a concept or measure when it comes to understanding how neighborhoods affect health. Some concerns include the validity of trust as a proxy for social capital (Carpiano & Fitterer, 2014), the conceptual difference between trust and mistrust (Wang, Schlesinger, Wang, & Hsiao, 2009), and the degree to which trust is stable over the life course and the context in which trust is enacted (Garoon et al., 2016). Hardin (2006) argued that trustworthiness is the basis for the more commonly applied concept of trust and that it is related to, but distinct from, other key concepts, such as social capital, reciprocity, and risk. Hardin suggested that familiarity is central to trustworthiness, because when we become familiar with entities, it provides us knowledge about how trustworthy they are or not. Many of our participants had lived over 20 years in their neighborhoods and as such had developed a high level of familiarity with those neighborhoods. However, in the face of rapid neighborhood change, core elements of neighborhood life may have become unfamiliar and less trustworthy—and through the temporal and other dimensions noted above, these effects could be amplified for older adults who may experience them more deeply and may have less sense of control over the situation. If placed within the context of cumulative inequality (Ferraro & Shippee, 2009), our findings suggest that, in addition to other disadvantages experienced across the life course, the most recent cohort of older African Americans in cities like Detroit are now at risk for stressful exposures at the neighborhood level. In many cases, neighborhoods that once may have buffered other forms of disadvantage, such as discrimination, have become untrustworthy places that now add additional burden in their lives.

Indeed, we discovered in our larger study sample that neighborhood disadvantage and perception of neighborhood quality were associated with higher salivary and hair cortisol levels (Zilioli, Slatcher, Fritz, Booza, & Cutchin, 2017). We therefore hypothesize that a fundamental pathway through which neighborhood stressors are experienced for older African Americans in United States “Rust Belt” cities is the multifaceted loss of trust in the neighborhood. Although an understudied issue in research about neighborhoods, stress, and health, trust may therefore provide a better starting point for conceptualizing how to address neighborhood level issues than the more widely used concept of social capital. Stress and health research in older populations might be informed by the dynamics of trust and trustworthiness.

We hasten to emphasize, however, that the role of neighborhood (as an ecology or place) should be central to such inquiry. An additional contribution of our analysis is how trust is part of the continuously evolving person–place relationship in later life. The insights provided by our data and analysis are therefore well aligned with current trends in environmental gerontology and geographical gerontology. While trust as a concept has been rarely used in those literatures, our emerging understanding of it from the Detroit data suggest a natural fit and contribution to those literatures. Increasingly relational, dynamic, and integrated frameworks of understanding environments in aging (e.g., Moore, 2014), and geographical gerontology’s insights into places, spaces, and landscapes of human experience and well-being (e.g., Skinner, Andrews, & Cutchin, 2018), are well poised to take such understandings and expand them.

Funding

This work was supported by funds from the Wayne State University Office of Vice President for Research and by the Wayne State University Center for Urban Responses to Environmental Stressors (CURES) (NIEHS Grant P30 ES020957).

Conflict of Interest

None reported.

Author Contributions

All of the authors have contributed to the final version of this manuscript and the contributions are documented below. H. Fritz was involved in the study design, data collection, qualitative analysis and interpretation of the findings, critical revision of all versions of the manuscript, and approval of the final version of the manuscript. E. R. Cummins was involved in data collection, qualitative analysis and interpretation of the findings, critical revision of all versions of the manuscript, and approval of the final version of the manuscript. M. P. Cutchin was involved in the study design, data collection, qualitative analysis and interpretation of the data, critical revision of all versions of the manuscript, and approval of the final version of the manuscript.