Evaluating the Relationship of Social Service Usage to Neighborhood Community for Older Adults: A Case Study

Abstract

Older adults who live in cities can cultivate a sense of neighborhood community that mitigates problems like poverty and ill health. However, it is not certain residence in urban neighborhoods is always related to community. This study investigates the potential of social services to positively associate with neighborhood community. Offering chances for older adults to interact with each other and with service providers who render basic services like social activities, meals, and housing, social services are a unique source of such community. We utilized hierarchical linear models with individual data from the Public Health Management Corporation’s Southeastern Pennsylvania Household Health Survey and neighborhood data from the American Community Survey. Findings suggest that the association with neighborhood community depends on the type of service and that race/ethnicity has an impact. For older adult Black respondents, residence in mostly Black neighborhoods increases the chances some services will relate to neighborhood community.

Keywords

neighborhood community older adults hierarchical linear models Public Health Management Corporation Philadelphia

Introduction

Older adults are one of the more vulnerable populations in the United States. They are more at risk to illnesses than younger adults (Ortman, Velkoff, and Hogan 2014) and experience greater impact from place-based problems like poverty (Klinenberg 2003; Park et al. 2017; Walker and Hiller 2007). Feeling connected to others outside one’s home may counteract these problems (Klinenberg 2003; Versey 2018), and urban neighborhoods can offer a sense of community that increases well-being (Sampson 2012; Small 2004). A strong sense of neighborhood community is related to a feeling of belonging as well as resources, thanks to neighbors working together to accomplish goals (McMillan and Chavis 1986). Evidence has shown that older adults with strong neighborhood community can more effectively manage disruptive events, like the 1995 Chicago heatwave (Klinenberg 2003).

Receiving social services in their neighborhood is one way for older adults to build neighborhood community (Sampson 2012). Older adult populations often rely on various social services (Cramm, van Dijk, and Nieboer 2013; Forsman et al. 2013; Lager, van Hoven, and Huigen 2015; Marwell 2007; Park et al. 2017; van Hoven and Douma 2012), which offers a method to build neighborhood community even in the absence of family or friends. However, research has not addressed at least two lingering questions about the nature of the association of social services and neighborhood community for older adults. First, what kind of services is most conducive to neighborhood community? While some services, like social activities, offer a well-defined space for social interaction that could result in community building (Putnam 2000), other services, like prescription support, have less opportunity for structured interaction. Second, does the neighborhood in which an older adult resides affect the impact of service usage on neighborhood community?

The questions guiding this article are thus twofold: Which services are more likely related to neighborhood community than others? and What neighborhood characteristics are more conducive to the relationship of services to neighborhood community? To address these questions, we make use of hierarchical linear modeling (HLM), which allows us to nest individual respondents into their respective neighborhoods to assess whether neighborhood context is associated with the variation in their sense of neighborhood community. Such a method allows us to better unpack the effects of residing in a neighborhood on an older adult’s neighborhood community (Gibbons and Yang 2016). Using the 2010, 2012, and 2014 waves of the Public Health Management Corporation’s (PHMC) Southeastern Pennsylvania Household Health Survey, we measure neighborhood community in two different ways: a sense of cooperation with neighbors and a sense of belongingness to one’s neighborhood. Both measures present different mechanisms through which neighborhood community can be cultivated and as such are useful to distinctly measure. The subsequent sections are organized as follows: The next section reviews the relation of social services neighborhood connectivity and proposes the research hypotheses of this study. We then discuss data, measures, and methods used in this study. The analytic results, discussion, and conclusions follow.

Background

Neighborhood community is the sense that members belong to a place as well as confidence in the ability of neighbors to cooperate to meet each other’s needs (McMillan and Chavis 1986). A sense of belongingness and confidence in local cooperation are distinct ways to feel attached to one’s local community, the former based upon a sense of local identity and the latter on the ability of a community to work together. Each has unique connotations for older adults, who may lack neighborhood community otherwise due to an absence of a strong local presence of family and friends (Ortman, Velkoff, and Hogan 2014).

One’s sense of cooperation with their neighbors means the sense of being able to work together with peers to accomplish good in one’s neighborhood, and such good might include assisting local older adults (Browning and Cagney 2002; Klinenberg 2003). The good from cooperation is especially important for older adults who live alone and are less mobile, in need of assistance with regular day-to-day activities (Fukui and Menjívar 2015; Klinenberg 2003; Perissinotto and Covinsky 2014; Versey 2018). A high level of cooperation is also a proxy of trust as it indicates faith in one’s neighborhood (Sampson 2012). These findings suggest that older adults who have a strong sense their neighborhood is cooperative feel the neighborhood is there to help them.

While one’s sense of belongingness to their neighborhood does not have the same connotation of social support associated with cooperation, it is also important for neighborhood community. The sense of local belongingness is extensively discussed in the literature on local attachment (Lewicka 2010; Scannell and Gifford 2010, 2017). Like cooperation, belongingness is a proxy of trust, as feeling like a part of a neighborhood reflects trust. Indeed, older adults who feel they belong to their neighborhoods are more likely to have self-esteem and a sense of meaning in life, which can improve their overall well-being (Buffel, De Donder, Phillipson, Dury, et al. 2014; Scannell and Gifford 2017).

Belongingness and a sense of cooperation are distinct indicators of neighborhood community; the first is directly complementary to a sense of attachment to a place and the second is faith in a neighborhood’s ability to get things done, the greater the faith the greater the neighborhood community. One can have belongingness without a sense of cooperation—which is common in low-income communities (Sampson 2012)—or vice versa. It is useful to consider belongingness and cooperation distinctly when determining the relation of service usage to neighborhood community.

Existing research has not examined the relationship between belongingness or cooperation and service usage. Arguably, the availability of services in a neighborhood suggests a neighborhood can cooperate for the common good. Likewise, receiving services motivates older adults to cooperate with one another (Chung 2007; Marwell 2007). This cooperation could help older adults develop a sense of neighborhood community. In addition, receiving local services promotes neighborhood belongingness as residents have more affinity for places which they feel support them (Buffel, De Donder, Phillipson, De Witte, et al. 2014; van Hoven and Douma 2012).

Local services have been shown to produce neighborhood community (Sampson 2012). Social services provide older adults opportunities for informal encounters with both service providers and, depending on the service, their peers who are also seeking out services (Cramm, van Dijk, and Nieboer 2013; Forsman et al. 2013; Lager, van Hoven, and Huigen 2015; van Hoven and Douma 2012). Thus, social services contribute toward the connectivity of older adults (Forsman et al. 2013; Walker and Hiller 2007). However, past work does not clarify which kind of services support neighborhood community most effectively. We know, for example, that older adults must perceive services to be age-friendly or they will not use them (Buffel, De Donder, Phillipson, Dury, et al. 2014). However, even older adult–focused services may not contribute toward neighborhood community. Park et al. (2017), for example, found that older adult housing services had little association with a local sense of connection. This could be explained by older adults being new to a neighborhood and unconnected to that place—however, the eventual stability in housing from these services would result in stability and by extension a greater sense of connectiveness (Oishi 2010; Sampson 2012). There is little existing research to establish which type of service matters, although Small’s (2009) research on childcare centers suggest that services that involve heavy social interaction are the most effective in building connectiveness.

Another potential influence on the association of services to neighborhood community is the character of the neighborhood itself. One relevant characteristic is organizational density. The largest suppliers of social services to urban neighborhoods are community-based organizations (CBOs; Allard 2009). CBOs play an important role in the association of services to neighborhood community. First, an organization can put a face on services that older adults are receiving. If the organization has a close affinity with a neighborhood, this can build a sense of neighborhood community for the residents within that neighborhood (Marwell 2007). Second, many local organizations use service provision as an opportunity to mobilize their local communities around local issues of importance, including community well-being and development. This civic engagement has the effect of further increasing residents’ sense of neighborhood community (Chung 2007; Marwell 2007; Sampson 2012).

On the contrary, research indicates limitations in CBOs’ ability to build neighborhood community. An organization’s mere presence in a neighborhood does not guarantee that it will be closely connected to that place or its residents (Kissane 2010). For example, Sampson (2012) found little association between organizational density and neighborhood community. However, Sampson did not distinguish respondents by age. The lower mobility of older adults than other residents may change the association, given that it makes them more dependent on nearby local organizations. At the same time, while CBOs are arguably the most significant service provider in urban neighborhoods (Allard 2009), services can come from a number of other sources, including less formal groups like block organizations and tenant associations (Sampson 2012), which lower older adults’ dependence on them and may lessen their impact, even for this group.

An additional neighborhood-level influence is racial/ethnic composition. Black residents who live in mostly Black neighborhoods have greater trust in health services (Ahern and Hendryx 2003; Gibbons 2019; Sewell 2015). This may be because they are more likely to receive care from service providers who are associated with the predominant race of their local neighborhoods (Gibbons 2015). In addition, people of color who live in predominantly non-White neighborhoods are less likely to feel service providers subject them to discrimination (Gibbons and Yang 2018). This heightened trust could enhance the potency of service providers in promoting neighborhood community for their Black older adult clients—if they live in predominantly Black (or at least non-White) neighborhoods. Indeed, older adults in neighborhoods where their own race is a minority often have less neighborhood community (Cummings 1998; Klinenberg 2003). However, there are limitations to the advantage of mostly non-White neighborhoods. Many predominantly non-White neighborhoods are a product of residential segregation, which is associated with systemic disadvantage which can erode neighborhood community (Sampson 2012). For example, Black older adults in low-income mostly Black Chicago neighborhoods were disproportionately affected by the 1995 Chicago heatwave as a result of the weakened community found in their neighborhoods (Klinenberg 2003).

A third neighborhood characteristic important to consider is socioeconomic disadvantage. The relationship of disadvantage to neighborhood community is complex. The connectivity of older adults in low-income neighborhoods is particularly vulnerable due to the perceived lack of cooperation found in these neighborhoods (Klinenberg 2003). Scholarship focused on older adults suggests that neighborhood poverty can undermine the role of services on neighborhood community (Buffel, De Donder, Phillipson, De Witte, et al. 2014). Furthermore, deeply impoverished neighborhoods often lack a strong nonprofit infrastructure, which can further reduce the chance for neighborhood community (Allard 2009). However, one study found that service-providing organizations can increase a sense of local civic identity in highly impoverished neighborhoods (Swaroop and Morenoff 2006).

In sum, there is ample evidence that social services can matter for the neighborhood community of older adults. Social service usage can positively relate to neighborhood community through several paths, including indirectly as a side effect of the interactions between clients and their service providers and other clients (Cramm, van Dijk, and Nieboer 2013; Forsman et al. 2013; Lager, van Hoven, and Huigen 2015; van Hoven and Douma 2012), and more directly as a result of service providers actively trying to facilitate a local civic identity among their clientele (Chung 2007; Marwell 2007). However, questions remain as to the nature of this association. First, which types of services have the greatest impact on older adult neighborhood community? It seems likely that services with a strong social component are the most likely to be associated with neighborhood community, but this has not been confirmed. Second, do services matter differently for neighborhood community depending on the aspect of neighborhood community measured, belongingness or cooperation? Past research has not provided a clear answer to this question. Third, how does the framing of neighborhood community affect the measured influence of services? For example, do social activities matter more for a sense of neighborhood belongingness than a sense of neighborhood cooperation? Also, there are several questions as for how the neighborhood an older adult resides in matters. To this end, we investigate, fourth, how important is the density of organizations in a neighborhood for the association of services to neighborhood community? Do services mainly influence neighborhoods that have a strong presence of CBOs, or does the role of social services exist independently of the local presence of CBOs? Fifth, how do older adults’ race and neighborhood racial/ethnic composition matter for neighborhood community? Are services more likely to associate with neighborhood belongingness for Blacks who live in mostly Black neighborhoods?

To investigate these questions, we explore the following hypotheses. First, we hypothesize that social service usage is related to neighborhood community, testing both social services positively relate to residents’ sense of neighborhood belongingness (Hypothesis 1 [H1]) and social services positively relate to residents’ sense of neighborhood cooperation (Hypothesis 2 [H2]). Next, based on our suspicion that social services with more explicit social dynamics, such as community activities, will be more related to neighborhood community, we hypothesize social services with a strong social dynamic will be more strongly associated with neighborhood community than other types of services (Hypothesis 3 [H3]). As for the question of organizational density, we hypothesize the presence of CBOs explain the relation of social services to neighborhood community (Hypothesis 4 [H4]). Finally, based upon the past research on race/ethnicity and services (Ahern and Hendryx 2003; Gibbons 2019; Sewell 2015), we suspect that local racial/ethnic composition has an important association with the relation of services to local connectivity. Thus, we examine services for Black residents living in mostly Black neighborhoods will have a stronger association with neighborhood community than those residing in other neighborhoods (Hypothesis 5 [H5]).

Data

To investigate our hypotheses, we utilize the 2010, 2012, and 2014/2015 waves of the PHMC’s Southeastern Pennsylvania Household Health Survey, a randomized telephone community survey of the Philadelphia metropolitan area. These were the most recently available waves of data with all the measures for which we were interested. Philadelphia is an ideal location to test our hypotheses as it has a robust social service infrastructure similar to many northeastern postindustrial cities (Allard 2009). PHMC surveys are reliable and a valid data sources for understanding health and socioeconomic status, with estimates similar to other surveys like the Behavioral Risk Factor Surveillance System (Gibbons and Yang 2016). The PHMC offers balancing weights that adjust for sampling bias and represent the population in the study area (PHMC 2016). It also provides geocodes that allow respondents to be nested in census tracts, a widely used proxy of neighborhoods for the quantitative study of neighborhood community (Gibbons and Yang 2016), easing the comparability of our results with other work. The response rate for the PHMC was 6.8% for the landline sample and 11.7% for the cell sample. While these rates are in keeping with response rates identified in other surveys like Pew (Kennedy and Hartig 2019), this may impact the generalizability of these results. Missing values, which accounted for 15% of the data set, were omitted through listwise deletion. The cleaned data set consists of 3,326 respondents residing in 377 tracts in the city of Philadelphia.

Individual-Level Variables

Our dependent variables are two measures of neighborhood community, Belongingness and Cooperation. Informed by existing theories of local community (McMillan and Chavis 1986), they are based upon the following survey questions which deal with neighborhood community: Belongingness: “Please tell me if you strongly agree, agree, disagree, or strongly disagree with the following statement: I feel that I belong and am a part of my neighborhood” (1 = strongly agree/agree, and 0 = disagree/strongly disagree), and Cooperation: “Have people in your neighborhood ever worked together to improve the neighborhood? For example, through a neighborhood watch, creating a community garden, building a community playground, or participating in a block party” (1 = yes, 0 = no). In developing these measures, the PHMC established their criterion and face validity (Harkins-Schwarz, Email Communicaiton to Author, April 14, 2020). In addition, these measures have been compared with other measures of neighborhood community and it was found they yield similar outcomes (Brinig and Garnett 2014).

As for our focal predictors, the PHMC has questions for several different types of services an older adult could receive, which are described in Table 1. Services with structured activities have a documented association with neighborhood community (Marwell 2007). Meanwhile, services with less assumed structured activities, like housing, have less of a documented association with neighborhood community (Park et al. 2017). Based on this literature, and the text of the PHMC questions, we speculate in Table 1 which services are the most likely to have a social association. We suspect that of all the services, “activities,” are most likely to support social interaction, given they directly deal with structured social activities. However, only our main analysis can determine for sure which services have more or less of an association with connectivity.

Table 1.

Neighborhood Typology.

Neighborhood Type	Neighborhood Characteristic
Predominantly White	Tracts where there are at least 60% of non-Hispanic Whites and no minority group represents more than 20%
Predominantly Black	Tracts that have at least 50% non-Hispanic Blacks and no more than 20% of another racial/ethnic group
Predominantly non-Black minority	Tracts consist of at least 50% of a certain non-Black minority and no more than 20% of non-Hispanic Blacks
Mixed	Tracts that cannot be classified into any of the typologies above

Our other predictors include indicators of race and ethnicity to account for racial and ethnic differences in neighborhood community (Gibbons and Yang 2016) classified into Black, Hispanic, and other non-Hispanic minorities. Also, we control for various measures of socioeconomic status given their direct link to one’s neighborhood community (Small 2004), including education attainment, classified as no high school (reference group), high school, and college educated or greater; marital status, categorized into single (reference group), married or living with a partner, widowed/divorced/separated (WDS), and another marital status; and living below the federal poverty line. Also, we account for the unique relation between those who live in owned or rented homes and neighborhood community with homeowner (1 = owner/0 = renter; Carson, Chappell, and Dujela 2010; Rohe and Stegman 1994). Furthermore, we account for older adult mobility with the instrumental activity of daily living (IADL) measure. Finally, we include the continuous measure of self-reported age, which is standardized in the multivariate analysis.

Our neighborhood measures include a measure of racial/ethnic composition based on Gibbons and Yang (2016) using 2010–2014 American Community Survey (ACS) data to determine how neighborhood racial/ethnic composition relates to community. This measure compares different compositions of non-Hispanic Whites, non-Hispanic Blacks, and “non-Black minorities,” including Hispanics and non-Hispanic Asians, and other racially mixed people. This typology allows us to contrast different segregated communities to assess the differential effects of residing in each for neighborhood community (Gibbons and Yang 2016). These classifications are discussed in Table 2. In our analysis, we use predominantly White neighborhoods as the reference group at the neighborhood level. Also, with the ACS data, we applied principal components analysis (PCA) to the following variables to create a measure of low socioeconomic status: female-headed household (0.859), unemployed (0.916), public assistance (0.872), less than high school (0.809), poverty (0.882), and married (–0.769). The resulting component accounted for 72% of the variation among these variables. Finally, we consider residential stability of a neighborhood given its previously recognized association with neighborhood community (Oishi 2010; Sampson 2012), averaging the standardized scores of the following two variables: percentage of owner-occupied housing units and percentage of residents who did not move for at least five years.

Table 2.

Breakdown of Services and Their Potential Sociability.

Service	Question Asked	Level of Sociability
Activities	Have you used activities programs at Senior Centers or Senior Clubs in the past year?	Most Likely Social
Housing	Have you used housing services such as energy assistance programs, subsidized housing, or home repair programs	Not Likely Social
Meals	Have you used meals/food programs such as home delivered meals, meals at Senior Centers or emergency food in the past year?	Potentially Social
PACE	Have you used PACE–Prescription Drug Program for the Elderly in the past year?	Not Likely Social
Transport	Have you used transportation services or resources (medical transportation, public transit discounts) in the past year?	Potentially Social

Note. PACE = Program of All-Inclusive Care for the Elderly.

Finally, we employ two measures of organizational density. First, we look at what Marwell and Gullickson (2013) identified as distributive organizations, CBOs that provide services within a city. Second, we made use of the City of Philadelphia’s Registered Community Organizations (RCOs). RCOs are nonprofit organizations that the city has designated to aid in local community issues. Unlike the distributive organizations, they have identified service boundaries. As such, we measure the number of distributive organizations located in a tract and the number of RCOs that have territory in that tract. Including both distributive organizations and RCOs offers a nuanced measure of organizational density, distinguishing organizations which have an explicit service area from organizations which may just have their headquarters in a neighborhood.

Several steps were taken to obtain and clean the organizational data. To identify distributive organizations, we followed the strategy of Marwell and Gullickson (2013), using Internal Revenue Service data packaged by the National Center for Charitable Statistics (NCCS) to produce a list of nonprofits with Philadelphia addresses. We then conducted a web search of each organization to determine whether they had a mission statement to provide services to Philadelphia communities. If an organization had no website, or their website did not specify their service area, we telephoned them to identify it (this took place in fall of 2015). Organizations that did not respond or did not have a local focus were not included in our data set. Distributive organizations were then geocoded using ArcGIS’s online services using their reported addresses. The RCO data were obtained from the OpenPhilly data portal. While most RCOs could be considered CBOs, some were specifically political organizations such as ward-level political parties. As these groups are not considered service providers in the same way as CBOs (Marwell 2007), we omitted them. There was some overlap between distributive organizations and RCOs. Distributive organizations that have RCO status are listed solely as RCOs to avoid organizational overcount. These efforts led to a total of 113 distributive organizations and 208 RCOs. While there are likely organizations missing from these counts, these data can be considered a workable proxy of what organizations are present in Philadelphia neighborhoods.

Results

We begin by describing our main predictors (Table 3). While most of the predictors are reported as proportions, for ease of interpretation, we discuss them here as percentages. First, 89.8% (0.898 × 100 = 89.8) of older adults feel they belong to their local neighborhoods. Cooperativeness was not quite as strong, as only 68.7% of residents feel their neighbors cooperate to address local issues. As for services, the most commonly used is transportation programs, with 20.3% of residents using them. The least commonly used are meal programs, which are used by only 8.7% of older adults. Turning to our other predictors, shares of Whites and Blacks in our sample are roughly similar, 49.96% of the respondents are White and 42.6% are Black. A large share of the sample is WDS, 47.1%, suggesting many live alone. Most of the respondents are retired (65.5%). Almost three quarters of the respondents, 74.8%, own their homes. Next, 29.8% of the sample reported needing help with at least one IADL, suggesting a lack of mobility and direct need of services. The organizational presence in a respondent’s neighborhood depends on the kind of organization. The average neighborhood has less than one distributive organization, 0.321, but is within the territory of six RCOs. Meanwhile, roughly 33.4% of the sample live in predominantly White neighborhoods while 38% live in predominantly Black neighborhoods. Given these numbers do not align with the racial character of the individuals in this sample, there is likely some mismatch between people’s race and the racial character of their neighborhoods.

Table 3.

Descriptives of Seniors in Philadelphia.

Statistic	M	SD
Individual level
Belonging	0.898	0.303
Cooperation	0.687	0.464
Black	0.426	0.495
White	0.496	0.500
Hispanic	0.048	0.214
Other	0.030	0.170
Age	70.775	8.171
Female	0.671	0.470
Married	0.312	0.464
Widowed/divorced/separated	0.471	0.499
Other marital status	0.003	0.057
Single	0.194	0.395
Poor	0.158	0.364
Full time	0.129	0.335
Part time	0.068	0.252
Retired	0.655	0.476
Other employment status	0.019	0.137
Unemployed	0.037	0.189
No high school	0.149	0.357
High school diploma	0.419	0.493
Some college	0.200	0.400
Bachelor’s degree	0.232	0.422
Homeowner	0.748	0.434
IADL	0.298	0.457
Housing	0.159	0.365
Meal	0.087	0.282
Transportation	0.203	0.402
PACE	0.115	0.319
Activity	0.190	0.392
Neighborhood level
SES	−0.015	2.098
RCO	6.004	3.507
Distributive organizations	0.321	0.695
Predominantly White	0.334	0.472
Predominantly Black	0.380	0.485
Non-Black minority neighborhood	0.038	0.191
Mixed neighborhood	0.249	0.432
Stability	−0.213	0.639
N	3,326

Note. IADL = instrumental activity of daily living; PACE = Program of All-Inclusive Care for the Elderly; SES = socioeconomic status; RCO = Registered Community Organizations.

To determine whether services are related to connectivity, we turn now to our logistic HLM, using the R package lme4. Our HLM models incorporate a random intercept but not a random slope. To ensure the appropriateness of the HLM approach, we first conducted a null model where no independent variable is included and determined whether the belongingness and cooperation differ by tracts, our proxy of neighborhoods. We find that both measures are significant, warranting an HLM approach. To determine the overall influence of tracts on these measures, we conduct intraclass correlation coefficient (ICC) tests. The variability of belongingness is relatively low; neighborhoods can explain only 3.33% of the variability of this measure. This falls outside of the minimum commonly accepted ICC score of 5% (Snijders and Bosker 1999). Much more variability for cooperation can be explained by tracts (13.42%).

We report our HLM results in Table 4 as odds ratios. We find that, as hypothesized, activity services are related to neighborhood community for older adults as measured by belongingness and cooperation. However, the significance of this association depends on the aspect of neighborhood community measured. While odds ratios for activity services are only marginally significant for belongingness (p < .100), they are fully significant for cooperation (p < .001). In short, those who use activity services have a 58.7% greater chance ([1.587 − 1] × 100) of feeling their neighborhood communities cooperate. Interestingly, housing services show a similar trend to social activities—a marginally significant relation with belongingness and a fully significant association with cooperation. Those who use housing services have a 50.1% greater chance of feeling their neighbors cooperate. This is notable given we did not anticipate sociability with housing services. Meanwhile, meal program recipients are 57.9% less likely to feel their neighbors cooperate.

Table 4.

Odds Ratios of Social Services and Neighborhoods (N = 3,326).

	Belonging		Cooperation
	(1)	(2)	(3)	(4)
Individual
Race/ethnicity (Ref: White)
Black	1.878***	2.649***	3.865***	2.697***
Black	(0.208)	(0.268)	(0.148)	(0.188)
Hispanic	1.377	1.741*	0.568**	0.462***
Hispanic	(0.312)	(0.330)	(0.222)	(0.243)
Other	1.553	1.983	1.035	0.856
Other	(0.476)	(0.487)	(0.303)	(0.307)
Age^a	1.024*	1.022*	0.977***	0.981**
Age^a	(0.012)	(0.012)	(0.008)	(0.008)
Female (Ref: Male)	0.763	0.786	0.814*	0.830
Female (Ref: Male)	(0.174)	(0.174)	(0.115)	(0.115)
Marital status (Ref: Single)
Married	1.559**	1.422	1.519***	1.518***
Married	(0.221)	(0.222)	(0.154)	(0.154)
Widowed/divorced/separated	1.077	0.999	1.842***	1.776***
Widowed/divorced/separated	(0.214)	(0.214)	(0.158)	(0.158)
Other marital status	0.733	0.597	3.764	4.199
Other marital status	(1.332)	(1.343)	(1.280)	(1.316)
Living in poverty	0.961	0.944	0.841	0.848
Living in poverty	(0.230)	(0.228)	(0.168)	(0.168)
Employment status (Ref: Unemployed)
Full time	1.491	1.593	1.107	1.138
Full time	(0.314)	(0.314)	(0.227)	(0.226)
Part time	1.812	1.888*	1.853**	1.864**
Part time	(0.380)	(0.382)	(0.274)	(0.273)
Retired	1.476	1.550*	0.834	0.824
Retired	(0.242)	(0.239)	(0.180)	(0.179)
Other employment status	0.522	0.555	0.951	0.965
Other employment status	(0.437)	(0.438)	(0.384)	(0.384)
Educational attainment (Ref: No high school degree)
High school diploma	1.358	1.340	1.216	1.273
High school diploma	(0.229)	(0.229)	(0.166)	(0.165)
Some college	1.139	1.084	1.298	1.322
Some college	(0.265)	(0.265)	(0.192)	(0.192)
Bachelor’s degree	2.469***	2.277***	2.617***	2.323***
Bachelor’s degree	(0.297)	(0.300)	(0.203)	(0.205)
Homeowner	1.514**	1.594**	1.514***	1.520***
Homeowner	(0.196)	(0.199)	(0.142)	(0.142)
IADL	0.801	0.851	0.933	0.972
IADL	(0.184)	(0.185)	(0.128)	(0.128)
Services
Housing	1.597*	1.574*	1.501**	1.556**
Housing	(0.263)	(0.261)	(0.186)	(0.186)
Meal	0.902	0.927	0.579***	0.590**
Meal	(0.308)	(0.306)	(0.211)	(0.211)
Transportation	1.097	1.097	1.244	1.156
Transportation	(0.226)	(0.225)	(0.155)	(0.155)
PACE	1.241	1.305	0.920	0.916
PACE	(0.297)	(0.296)	(0.190)	(0.190)
Activities	1.627*	1.706**	1.587***	1.577***
Activities	(0.253)	(0.254)	(0.159)	(0.157)
Neighborhood
Racial composition (Ref: White)
Predominantly Black		0.446**		2.388***
Predominantly Black		(0.378)		(0.266)
Non-Black minority		0.661		4.060***
Non-Black minority		(0.589)		(0.431)
Mixed neighborhood		0.445***		0.983
Mixed neighborhood		(0.295)		(0.199)
Socioeconomic status		0.943		0.827***
Socioeconomic status		(0.073)		(0.053)
RCO		1.047		1.052**
RCO		(0.033)		(0.023)
Distributive organization		0.955		1.086
Distributive organization		(0.157)		(0.107)
Stability		1.131		0.738**
Stability		(0.173)		(0.123)
Constant	0.657	0.802	3.167**	1.294
Constant	(0.863)	(0.893)	(0.584)	(0.599)
Log likelihood	−677.631	−669.477	−1,224.765	−1,206.122

Note. Robust standard errors are given in parentheses. IADL = instrumental activity of daily living; PACE = Program of All-Inclusive Care for the Elderly; RCO = Registered Community Organizations.

p < .1.*p < .05.**p < .01.

Turning to our other predictors, Black older adults have a 87.8% greater chance than White older adults of feeling they belong to their neighborhoods and a 286.5% greater chance of feeling their neighbors cooperate. Next, older adults who are married have a 55.9% greater chance of feeling they belong to their neighborhoods and a 51.9% greater chance of feeling their neighbors cooperate than those who are single. Education also matters for community association: Those with a bachelor’s degree or greater have a 146.9% greater chance of neighborhood belongingness and a 161.7% greater chance of cooperation than those with no high school degree. Last, homeowners have a 51.4% greater chance of feeling both belongingness and cooperation.

Adding in neighborhood predictors has little effect on any of the service outcomes; most of the odds ratios for the individual-level predictors are largely similar in Models 2 and 4 as in Models 1 and 2. The neighborhood predictors themselves have a mixed relation with neighborhood community. For example, while residence in predominantly Black neighborhoods is related to 55.4% less of a chance of feeling belongingness, it is also related to a 138.8% greater chance of feeling neighbors cooperate. Next, each point increase in poor neighborhood socioeconomic status is related to a 17.3% decline in the chance one feels their neighbors cooperate. Last, with each addition of RCOs in a neighborhood, the chance one feels their neighbors cooperate increases by 5.2%. Distributive organizations are not significant.

To examine our fourth hypothesis, we conduct additional analysis with models using only the Black older adults in the sample, reported in Table 5. First, unlike the overall sample, using activity services is not significantly related to belongingness for Black older adults, even with the inclusion of neighborhood controls. Activities are only fully significantly related to cooperation for Black older adults with the inclusion of neighborhood controls, suggesting an association with neighborhoods, which will be explored in greater detail shortly. No other service is related to neighborhood community for the Black older adult sample. Some of the other predictors indicate similar associations with belongingness found in the models reported in Table 4. For example, marital status is also important for cooperation for Black older adults. Other predictors, however, diverge in their associations. For example, while, in the overall sample, a bachelor’s degree and homeownership were significantly associated with both belongingness and cooperation, for the Black sample, a bachelor’s degree is only related to cooperation and homeownership is only related to belongingness. Most notable was that residing in Black neighborhoods did not have an association with neighborhood community for the Black older adults.

Table 5.

Odds Ratios of Social Services and Neighborhoods for Blacks Only (N = 1,418).

	Belonging			Cooperation
	(5)	(6)	(7)	(8)	(7)	(8)
Individual
Age	1.026	1.025	1.035	0.986	0.989	0.987
Age	(0.021)	(0.021)	(0.022)	(0.014)	(0.014)	(0.014)
Female (Ref: Male)	0.836	0.742	0.772	0.632**	0.634**	0.643*
Female (Ref: Male)	(0.302)	(0.307)	(0.311)	(0.223)	(0.225)	(0.227)
Marital status (Ref: Single)
Married	1.716	2.008*	2.014*	2.205***	1.890**	2.118***
Married	(0.396)	(0.403)	(0.405)	(0.277)	(0.282)	(0.284)
Widowed/divorced/separated	1.011	1.081	1.014	2.542***	2.118***	2.311***
Widowed/divorced/separated	(0.339)	(0.340)	(0.345)	(0.263)	(0.262)	(0.265)
Other marital status	4.485	7.310	37.520	0.989	0.848	0.630
Other marital status	(4.202)	(5.520)	(12.936)	(1.875)	(1.848)	(1.871)
Living in poverty	1.143	1.346	1.241	1.184	1.074	1.165
Living in poverty	(0.350)	(0.354)	(0.357)	(0.272)	(0.274)	(0.280)
Employment status (Ref: Unemployed)
Full time	1.672	1.999	2.052	1.536	1.639	1.862
Full time	(0.520)	(0.533)	(0.540)	(0.437)	(0.442)	(0.446)
Part time	1.873	1.986	1.732	2.786	2.315	2.319
Part time	(0.691)	(0.702)	(0.689)	(0.660)	(0.653)	(0.651)
Retired	1.759	1.784	1.702	0.768	0.786	0.827
Retired	(0.376)	(0.380)	(0.383)	(0.304)	(0.303)	(0.304)
Other employment status	279.375	436.839	313.853	2.934	2.765	2.267
Other employment status	(8.382)	(10.600)	(10.343)	(1.114)	(1.099)	(1.071)
Educational attainment (Ref: No high school degree)
High school diploma	1.715	1.811	1.585	1.553	1.540	1.574
High school diploma	(0.378)	(0.382)	(0.391)	(0.282)	(0.281)	(0.283)
Some college	1.120	1.300	1.119	1.494	1.455	1.463
Some college	(0.428)	(0.443)	(0.448)	(0.336)	(0.340)	(0.343)
Bachelor’s degree	1.945	2.190	2.127	3.229***	3.045**	3.149***
Bachelor’s degree	(0.561)	(0.575)	(0.586)	(0.435)	(0.440)	(0.442)
Homeowner	3.453***	3.333***	3.022***	1.489*	1.471	1.402
Homeowner	(0.310)	(0.311)	(0.313)	(0.238)	(0.239)	(0.239)
IADL	0.861	0.882	0.843	1.244	1.286	1.370
IADL	(0.312)	(0.315)	(0.316)	(0.236)	(0.238)	(0.239)
Services
Housing	1.721	1.614	1.385	1.450	1.477	1.382
Housing	(0.392)	(0.393)	(0.828)	(0.292)	(0.297)	(0.538)
Meal	1.669	1.725	0.893	0.838	0.835	0.764
Meal	(0.610)	(0.614)	(1.036)	(0.384)	(0.381)	(0.741)
Transportation	2.076*	2.218*	1.978	0.985	1.026	1.044
Transportation	(0.403)	(0.417)	(0.425)	(0.258)	(0.260)	(0.263)
PACE	1.175	1.112	0.675	1.261	1.185	2.461
PACE	(0.565)	(0.561)	(1.076)	(0.381)	(0.387)	(0.885)
Activities	0.872	0.874	5.682	1.808*	1.905**	3.851**
Activities	(0.392)	(0.396)	(1.344)	(0.308)	(0.308)	(0.658)
Neighborhood
Racial composition (Ref: White)
Predominantly Black		3.034	3.064		1.933	2.184
Predominantly Black		(0.749)	(0.774)		(0.706)	(0.708)
Non-Black minority		4.277	3.923		4.540	4.420
Non-Black minority		(1.225)	(1.229)		(1.005)	(1.005)
Mixed neighborhood		5.289**	5.074**		0.988	0.988
Mixed neighborhood		(0.731)	(0.753)		(0.692)	(0.690)
Socioeconomic status		1.022	1.040		0.823**	0.835**
Socioeconomic status		(0.110)	(0.111)		(0.080)	(0.080)
RCO		1.014	1.022		1.062*	1.065*
RCO		(0.043)	(0.044)		(0.032)	(0.032)
Distributive organization		0.857	0.803		0.770*	0.752*
Distributive organization		(0.222)	(0.224)		(0.148)	(0.148)
Stability		1.206	1.283		1.048	1.059
Stability		(0.270)	(0.275)		(0.208)	(0.209)
Interactions
Activities × Predominantly Black			0.103			0.386
Activities × Predominantly Black			(1.409)			(0.743)
PACE × Predominantly Black			1.805			0.406
PACE × Predominantly Black			(1.266)			(0.979)
Meal × Predominantly Black			2.863			1.199
Meal × Predominantly Black			(1.284)			(0.856)
Housing × Predominantly Black			1.239			1.067
Housing × Predominantly Black			(0.919)			(0.614)
Constant	0.369	0.096	0.064	4.074	2.027	1.781
Constant	(1.477)	(1.685)	(1.715)	(1.028)	(1.238)	(1.228)
Log likelihood	−217.321	−213.380	−211.046	−358.580	−349.355	−347.898

Note. Robust standard errors are given in parentheses. IADL = instrumental activity of daily living; PACE = Program of All-Inclusive Care for the Elderly; RCO = Registered Community Organizations.

p < .1. **p < .05. ***p < .01.

We also included cross-level interactions of predominantly Black neighborhoods to services to see whether a Black respondent’s neighborhood help explains the association of services to neighborhood community. We find none of the interaction terms are significant. Their inclusion also has little effect on the odds ratios for belongingness, except that the transportation odds ratio lost significance. The most notable change in these models is the odds ratio for activities and cooperation, which is much higher in magnitude with the inclusion of the interaction terms. The odds ratio goes from 90.1% in the two-level model to 285.1% in the cross-level model. This suggests that the association of activity services with cooperation is stronger for Black older adults residing in predominantly Black neighborhoods.

Discussion

Social services offer the potential to build a sense of neighborhood community for older adults. However, there is much we do not know about this association. Two questions guided this study’s attempt to address this gap: Which services are more likely related to neighborhood community than others? and What neighborhood characteristics are more conducive to the relationship of services to neighborhood community? To address these questions, we evaluated two forms of neighborhood community: respondents’ sense of belongingness to their neighborhood and their sense they can cooperate with their neighbors.

Overall, the results support our initial suspicions that not all services would have the same association with neighborhood community. However, some of the associations deviated from our expectations. We find first that some social services are indeed positively related to cooperation and belongingness. This includes social activities and housing services, which are both significantly related to cooperation and marginally related to belongingness. The association of activities and belongingness becomes fully significant with the neighborhood controls. This offers support for our first and second hypotheses: social services positively relate to one’s sense of neighborhood belongingness (H1) and social services positively relate to one’s sense of neighborhood cooperation (H2). These findings have a few important implications. For one, they offer empirical evidence that social services are related to older adults feeling that they belong to a place and that people look out for one another. The difference in significance of the associations of social activities to cooperation and belongingness is also important, demonstrating that services do not have the same degree of association with different forms of neighborhood community; the relation of activities to cooperation being more robust than activities to belongingness. Also, the findings on activity services offer some support for our third hypothesis, social services with a strong social dynamic will be more associated with neighborhood community than other types of services (H3), but we did not get full support for this hypothesis as we found housing services also had a relation to neighborhood community.

The association of housing services to neighborhood was unexpected. Our impressions from the phrasing of the survey question, as well as existing literature (Park et al. 2017), were that housing services did not have a social component. We suspect that housing services has more of an indirect social effect than activity services, because they do not have a strong interactive component. Instead, we speculate that housing services are creating the circumstances that facilitate neighborhood community, stable housing (Oishi 2010; Sampson 2012). Future research might probe this association. It is also worth noting housing services had no association with neighborhood community for the Black population, suggesting this benefit is race/ethnicity specific. More research should be done to explore this association as well.

Next, meal programs did not support neighborhood cooperation. The wording of the PHMC survey, reported in Table 2, would suggest meal programs have some degree of social interaction that would lead to neighborhood community. However, meal programs are not associated with belongingness or cooperation for any group, and have a negative association with cooperation for non-Black older adults. Future research should examine this finding and what may cause it.

As for our second research question, does neighborhood context relate to the association of services to neighborhood community, the evidence is more mixed. For one, the introduction of the neighborhood-level controls in our models, including racial composition, socioeconomic status, and organizational density, has little impact on the existing associations of services with neighborhood community. The only deviation from this trend is the association of activities to belonging, which becomes fully significant with the inclusion of the neighborhood-level controls.

The lack of notable change in the association of services to neighborhood community when controlling for neighborhood characteristics implies our fourth hypothesis, relation of social services to neighborhood community can be explained by the presence of CBOs (H4), is not supported. This finding supports past research showing the local density of organizations within a neighborhood does not matter for the relationship of services to neighborhood community (Kissane 2010; Sampson 2012). On the contrary, living in RCO territory is related to a sense of cooperation. This supports past research that locally focused CBOs can cultivate a local sense of cooperation in the places they work (Marwell 2007; Sampson 2012). Also, the fact that distributive organizations are not associated with either form of neighborhood community suggests they only support community if they have an explicit local focus.

The lack of significance of socioeconomic status for non-Black populations suggests social services can boost an older adult’s sense of neighborhood community, regardless of the disadvantage of the neighborhood where they reside. This finding amplifies our understanding of mixed results of past research on neighborhood policy and the association between neighborhood community and social services (Buffel, De Donder, Phillipson, De Witte, et al. 2014; Swaroop and Morenoff 2006). The finding suggests that social services can act as a measure to supplement the neighborhood community of older adults who live in disadvantaged neighborhoods. This benefit could prove to be essential for older adults who are vulnerable to problems from living in low-income neighborhoods (Klinenberg 2003). However, in the Black-only sample, disadvantage is negatively related to cooperation, suggesting social services alone might not provide a benefit.

However, closer consideration reveals that the impact of social services for Black respondents varies by neighborhood racial/ethnic composition. While predominantly Black neighborhoods have a negative relation with neighborhood community and positive association with cooperation, the Black alone models yield more meaningful findings. The inclusion of predominantly Black and the other neighborhood controls shows activity services have a significant positive association with cooperation. Moreover, the interaction terms suggest activity programs have a stronger relation to Black residents of mostly Black areas. This supports our fifth hypothesis: Services for Black residents living in mostly Black neighborhoods will have a stronger association with neighborhood community than those residing in other neighborhoods (H5). Our findings support the ample research that shows neighborhood race/ethnicity is related to neighborhood community for minority respondents. For example, Black older adults living in mostly Black neighborhoods have greater trust in service providers than those who live in other neighborhoods (Ahern and Hendryx 2003; Gibbons 2019; Sewell 2015), and we suspect this plays a role in the association of services to neighborhood community.

This research has a few limitations of note. First, as we do not have longitudinal data, we cannot verify a causal association between services and neighborhood community for individuals. Second, we do not know the specific source of the services reported in the data set. Consequently, we cannot evaluate whether services came from informal sources or more formalized CBOs. Moreover, we cannot say where exactly the services came from. As such, we do not know the impact of whether important services originate in residents’ own neighborhood. Third, Philadelphia’s high segregation and racial/ethnic composition may have an impact on outcomes that would not be present in other cities. Similar studies in other cities with different levels of segregation could provide further illumination.

Ensuring older adults have a sense of neighborhood community is important for their well-being. This study demonstrates that some types of social services address this critical need. This reiterates a long-held view that social services provide more than their direct, stated impact on the individuals who participate. Indeed, the findings of this study suggest the informal community benefit they can create adds enormously to their value.

Footnotes

Acknowledgements

Thanks to Liza Shifrin and Le Sony’r Ra for their research assistance and Kate Epstein and the anonymous reviewers for their instructive comments in developing this manuscript.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Joseph Gibbons

Author Biography

Joseph Gibbons is an associate professor of sociology and associate director at the Center for Human Dynamics in the Mobile Age at San Diego State University. His substantive research interests emphasize the implications of gentrification and racial segregation as they relate to health, neighborhood community, and residential attainment. He is currently developing a study on the ability of communities to socially distance in response to the COVID-19 pandemic.

References

Ahern

Melissa M.

Hendryx

Michael S.

2003. “Social Capital and Trust in Providers.” Social Science & Medicine 57 (7): 1195–203. doi:10.1016/S0277-9536(02)00494-X.

Allard

Scott W.

2009. Out of Place: The New Geography of Welfare Policy. Chicago: The University of Chicago Press.

Brinig

Margaret F.

Garnett

Nicole Stelle

. 2014. Lost Classroom, Lost Community: Catholic Schools’ Importance in Urban America. Chicago: The University of Chicago Press.

Browning

Christopher R.

Cagney

Kathleen A.

2002. “Neighborhood Structural Disadvantage, Collective Efficacy, and Self-Rated Physical Health in an Urban Setting.” Journal of Health and Social Behavior 43 (4): 383–99. doi:10.2307/3090233.

Buffel

Tine

De Donder

Liesbeth

Phillipson

Chris

De Witte

Nico

Dury

Sarah

Verté

Dominique

. 2014. “Place Attachment among Older Adults Living in Four Communities in Flanders, Belgium.” Housing Studies 29 (6): 800–22. doi:10.1080/02673037.2014.898741.

Buffel

Tine

De Donder

Liesbeth

Phillipson

Chris

Dury

Sarah

De Witte

Nico

Verté

Dominique

. 2014. “Social Participation among Older Adults Living in Medium-Sized Cities in Belgium: The Role of Neighbourhood Perceptions.” Health Promotion International 29 (4): 655–68. doi:10.1093/heapro/dat009.

Carson

Arlene J.

Chappell

Neena L.

Dujela

Carren E.

2010. “Power Dynamics and Perceptions of Neighbourhood Attachment and Involvement: Effects of Length of Residency versus Home Ownership.” Housing, Theory and Society 27 (2): 162–77. doi:10.1080/14036090903159960.

Chung

Angie Y.

2007. Legacies of Struggle: Conflict and Cooperation in Korean American Politics. Palo Alto: Stanford University Press.

Cramm

Jane M.

van Dijk

Hanna M.

Nieboer

Anna P.

2013. “The Importance of Neighborhood Social Cohesion and Social Capital for the Well Being of Older Adults in the Community.” The Gerontologist 53 (1): 142–52. doi:10.1093/geront/gns052.

10.

Cummings

Scott

. 1998. Left behind in Rosedale: Race Relations and the Collapse of Community Institutions. CRC Press. https://www.crcpress.com/Left-Behind-In-Rosedale-Race-Relations-And-The-Collapse-Of-Community-Institutions/Cummings/p/book/9780813334219.

11.

Forsman

A. K., C. Herberts

Nyqvist

Fredrica

Wahlbeck

Kristian

Schierenbeck

Isabell

. 2013. “Understanding the Role of Social Capital for Mental Wellbeing among Older Adults.” Ageing & Society 33 (5): 804–25. doi:10.1017/S0144686X12000256.

12.

Fukui

Haruna M.

Menjívar

Cecilia

. 2015. “Bound by Inequality: The Social Capital of Older Asians and Latinos in Phoenix, Arizona.” Ethnography 16 (4): 416–37.

13.

Gibbons

Joseph R.

2015. “Does Racial Segregation Make Community-Based Organizations More Territorial? Evidence from Newark, NJ, and Jersey City, NJ.” Journal of Urban Affairs 37 (5): 600–19. doi:10.1111/juaf.12170.

14.

Gibbons

Joseph R.

2019. “The Effect of Segregated Cities on Ethnoracial Minority Healthcare System Distrust.” City & Community 18 (1): 321–43. doi:10.1111/cico.12370.

15.

Gibbons

Joseph R.

Yang

Tse-Chuan

. 2016. “Connecting Across the Divides of Race/Ethnicity: How Does Segregation Matter?” Urban Affairs Review 52 (4): 531–58.

16.

Gibbons

Joseph R.

Yang

Tse-Chuan

. 2018. “Searching for Silver Linings: Is Perceived Medical Discrimination Weaker in Segregated Areas?” Applied Spatial Analysis and Policy 11 (1): 37–58. doi:10.1007/s12061-016-9211-5.

17.

Kennedy

Courtney

Hartig

Hanna

. 2019. “Response Rates in Telephone Surveys Have Resumed Their Decline.” Pew Research Center, February 27. https://www.pewresearch.org/fact-tank/2019/02/27/response-rates-in-telephone-surveys-have-resumed-their-decline/.

18.

Kissane

Rebecca Joyce

. 2010. “‘We Call It the Badlands’: How Social-Spatial Geographies Influence Social Service Use.” Social Service Review 84 (1): 3–28.

19.

Klinenberg

Eric

. 2003. Heat Wave: A Social Autopsy of Disaster in Chicago. Chicago: The University of Chicago Press.

20.

Lager

Debbie

van Hoven

Bettina

Huigen

Paulus P. P.

2015. “Understanding Older Adults’ Social Capital in Place: Obstacles to and Opportunities for Social Contacts in the Neighbourhood.” Geoforum 59 (February): 87–97. doi:10.1016/j.geoforum.2014.12.009.

21.

Lewicka

Maria

. 2010. “What Makes Neighborhood Different from Home and City? Effects of Place Scale on Place Attachment.” Journal of Environmental Psychology 30 (1): 35–51. doi:10.1016/j.jenvp.2009.05.004.

22.

Marwell

Nicole P.

2007. Bargaining for Brooklyn: Community Organizations in the Entrepreneurial City. Berkeley: University of California Press.

23.

Marwell

Nicole P.

Gullickson

Aaron

. 2013. “Inequality in the Spatial Allocation of Social Services: Government Contracts to Nonprofit Organizations in New York City.” Social Service Review 87 (2): 319–53. doi:10.1086/670910.

24.

McMillan

David W.

Chavis

David M.

1986. “Sense of Community: A Definition and Theory.” Journal of Community Psychology 14 (1): 6–23.

25.

Oishi

Shigehiro

. 2010. “The Psychology of Residential Mobility: Implications for the Self, Social Relationships, and Well-Being.” Perspectives on Psychological Science 5 (1): 5–21. doi:10.1177/1745691609356781.

26.

Ortman

Jennifer M.

Velkoff

Victoria A.

Hogan

Howard

. 2014. “An Aging Nation: The Older Population in the United States.” Washington, DC: U.S. Census Bureau.

27.

Park

Sojung

Han

Yoonsun

Kim

BoRin

Dunkle

Ruth E.

2017. “Aging in Place of Vulnerable Older Adults: Person–Environment Fit Perspective.” Journal of Applied Gerontology 36 (11): 1327–50. doi:10.1177/0733464815617286.

28.

Perissinotto

Carla M.

Covinsky

Kenneth E.

2014. “Living Alone, Socially Isolated or Lonely—What Are We Measuring?” Journal of General Internal Medicine 29 (11): 1429–31. doi:10.1007/s11606-014-2977-8.

29.

Public Health Management Corporation. 2016. “2014-2015 Household Health Survey Documentation.” Philadelphia: Public Health Management Corporation.

30.

Putnam

Robert D.

2000. Bowling Alone: The Collapse and Revival of American Community. New York: Simon & Schuster.

31.

Rohe

William M.

Stegman

Michael A.

1994. “The Impact of Home Ownership on the Social and Political Involvement of Low-Income People.” Urban Affairs Quarterly 30 (1): 152–72. doi:10.1177/004208169403000108.

32.

Sampson

Robert J.

2012. Great American City: Chicago and the Enduring Neighborhood Effect. Chicago: The University of Chicago Press.

33.

Scannell

Leila

Gifford

Robert

. 2010. “Defining Place Attachment: A Tripartite Organizing Framework.” Journal of Environmental Psychology 30 (1): 1–10. doi:10.1016/j.jenvp.2009.09.006.

34.

Scannell

Leila

Gifford

Robert

. 2017. “Place Attachment Enhances Psychological Need Satisfaction.” Environment and Behavior 49 (4): 359–89. doi:10.1177/0013916516637648.

35.

Sewell

Abigail A.

2015. “Disaggregating Ethnoracial Disparities in Physician Trust.” Social Science Research 54 (November): 1–20. doi:10.1016/j.ssresearch.2015.06.020.

36.

Small

Mario Luis

. 2004. Villa Victoria: The Transformation of Social Capital in a Boston Barrio. Chicago: The University of Chicago Press.

37.

Small

Mario Luis.

2009. Unanticipated Gains: Origins of Network Inequality in Everyday Life. New York: Oxford University Press.

38.

Snijders

Tom

Bosker

Roel

. 1999. Multilevel Analysis: An Introduction to Basic and Advanced Multilevel Modeling. Thousand Oaks: Sage.

39.

Swaroop

Sapna

Morenoff

Jeffrey D.

2006. “Building Community: The Neighborhood Context of Social Organization.” Social Forces 84 (3): 1665–95.

40.

van Hoven

Bettina

Douma

Linden

. 2012. “‘We Make Ourselves at Home Wherever We Are’—Older People’s Placemaking in Newton Hall.” European Spatial Research and Policy 19 (July): 65–79. doi:10.2478/v10105-012-0006-7.

41.

Versey

H. Shellae

. 2018. “A Tale of Two Harlems: Gentrification, Social Capital, and Implications for Aging in Place.” Social Science & Medicine 214 (October): 1–11. doi:10.1016/j.socscimed.2018.07.024.

42.

Walker

Ruth B.

Hiller

Janet E.

2007. “Places and Health: A Qualitative Study to Explore How Older Women Living Alone Perceive the Social and Physical Dimensions of Their Neighbourhoods.” Social Science & Medicine 65 (6): 1154–65. doi:10.1016/j.socscimed.2007.04.031.