Abstract
Recent decades have seen a growing number of studies on the benefits and environmental determinants of community social capital. This study explored the relationship between neighborhood residents’ perceptions of their built environment and social capital by comparing two neighborhoods, Bucktown, an example of traditional neighborhood design, and Schaumburg, exemplifying suburban sprawl. Furthermore, the study sought to develop suggestions for further research about the variables contributing to neighborhood variations in social capital. Results of two cross-sectional phone surveys with 197 residents indicated that Bucktown respondents reported more close neighborhood ties, and believed they were more involved in mutual aid and community problem solving, but viewed their neighbors as less supportive than participants from Schaumburg. It may be hypothesized that aspects of residents’ perceptions of the built environment, particularly perceived safety and walkability, may be partly responsible for the neighborhood differences found. Further research is needed to understand the pathways of how elements of perceived built environment may affect social capital formation and development.
Keywords
Since 1990s, social capital has emerged as a term of increasing popularity in research and social policy. In particular, recent research focuses on the benefits and environmental determinants of community social capital. Community social capital has been found to be associated with a variety of positive outcomes, including reduced rates of violent crime (Browning, Feinberg, & Dietz, 2004), firearm victimization (Medina, 2015), delinquency (Hagan, Merkens, & Boehnke, 1995; Roche, 1998), decreased fear of crime (Ferguson & Mindel, 2007; Kruger, Hutchison, Monroe, Reischl, & Morrel-Samuels, 2007), and increased neighborhood stability (Temkin & Rohe, 1998). High level of individual and/or community social capital may decrease the risk of a variety of health problems such as obesity (Muckenhuber, Dorner, Burkert, Groschädl, & Freidl, 2015) and lead to the improvement in individual and community health (De Silva, Huttly, Harpham, & Kenward, 2007; Kawachi, Kennedy, & Glass, 1999; Kawachi, Kennedy, Lochen, & Prothrow-Smith, 1997). Not surprisingly, social capital may have come to be viewed as a cure-all “panacea for maladies affecting society” (Aguilar & Sen, 2009, p. 424).
One should distinguish between the social capital as an individual good, or “the ability of actors to secure benefits by virtue of membership in social networks or other social structures” (Portes, 1998, cited in Eriksson & Emmelin, 2013, p. 113), and “social capital as community-level public good” (Briggs, 2004), “a collective feature that characterizes geographical areas in terms of levels of trust, reciprocity, and civic engagement” (Karhina, Ng, Ghazinour, & Eriksson, 2016, p. 2). Alternatively, following Lin (2001), social capital can be defined as resources derived from individual social networks. The present study focuses on neighborhood social capital and defines it as an individual good or resources of the residents of spatial communities or neighborhoods, conceptualizing social capital as residents’ neighborhood networks, norms, and trust that facilitate cooperation for mutual benefit (Putnam, 2000). In this context, social capital has four key dimensions: “sense of community, collective efficacy, neighboring, and participation” (Perkins & Long, 2002, cited in O’Connor, 2013, p. 973). Neighborhood social capital encompasses elements as diverse as socializing with neighbors (Guest & Wierzbicki, 1999), knowing neighbors and exchanging practical favors, friendships in the neighborhood, trust shared between neighbors, community cohesion, and overall quality of neighbor relationships (Völker & Flap, 2007). Generally, the emphasis here is in on the weak neighborhood ties, which, according to Sampson and associates (1997), are critical for enforcing the norms of informal social control and controlling social and physical disorder (Lelieveldt, 2004, cited in Völker & Flap, 2007).
Environmental Determinants of Social Capital
Since late 1990s, a growing number of studies have explored environmental factors that may affect the development of intra- or extra-neighborhood social capital networks (Briggs, 2004; Hutchinson, 2004; Payne & Williams, 2008; Saegert, Thompson, & Warren, 2001; Vidal, 2004). In particular, prominent scholarship in environmental psychology examines the effects of physical environment (both built and natural) on individuals and communities. Note that the physical environment includes the natural world and built world (Germain, 1979), and built environment “encompasses those aspects of the physical environment that are man-made—buildings, transportation systems, communication systems, recreational facilities, indeed the whole structure of material culture” (Kemp, Whittaker, & Tracy, 1997, p. 84). Environmental psychology sees behavior as partly a function of the place(s) or immediate physical environments where an individual spends time, with a special focus on how conducive these places are for one’s social interaction and participation (Saleebey, 2004). Broadly, Wood and Giles-Corti (2008) distinguish between three general domains in how physical environment can affect residents’ patterns of social interaction and social capital in a neighborhood. First, macrolevel environmental trends of crime, disorder, and residential instability can disrupt neighborhood ties, and undermine trust and neighborly interactions. Second, mezzo-level neighborhood characteristics can encourage or discourage walking and interaction, including street lighting, footpaths, and pedestrian-friendly street grid, and availability of local facilities in a neighborhood (e.g., shops, public transportation, etc.) within walking distance. In contrast, higher presence and speed of traffic, along with more parking on the street, can negatively affect elements of social capital (Muelle, 2003, as cited in Wood & Giles-Corti, 2008). Third, Wood and Giles-Corti (2008) emphasize microlevel or individual variables include the perceptions individuals have of their neighborhood environment—its safety, perceived opportunities for interaction and participation, perception of neighborhood disorder, and/or upkeep of public or private property—and perception of availability of open spaces and other aspects of neighborhood quality (local services, design, aesthetics, and area history).
Skjaeveland and Garling (1997) identify a number of objective characteristics of the built and natural environments that encourage social interaction in a community and improve residents’ perception of the neighborhood. Fundamentally, it is important to account for the functionality of a place, or availability of physical design elements or basic structures that can be used for the purposes of social interaction (e.g., benches, stairs, sidewalks, etc.; Skjaeveland & Garling, 1997). Similarly, presence of semiprivate spaces, or the in-between spaces between public and private “zones” (e.g., front gardens, porches, and verandas), will provide opportunities for interaction and increased surveillance in a neighborhood, enhancing visibility and residents’ sense of safety. The general appearance of the community—its attractiveness, or aesthetically pleasing design elements, and the architectural character of a neighborhood—is equally significant as it invites residents to venture out and “linger” longer (Skjaeveland & Garling, 1997). Another element of the built environment that may encourage informal neighboring is “spaciousness,” or residents’ perception of the openness of the community, along with some structuring of space (e.g., using edges, screens, or enclosures to enhance privacy). Similar to community aesthetics, spaciousness may enhance social interactions indirectly by improving residents’ perceptions of the neighborhood and inviting one to “linger.”
According to a number of authors, access to and presence or availability of spaces in a neighborhood that can serve as places to interact appears to be a critical component of the built environment that seeks to facilitate the formation of social capital. Such spaces include playgrounds, benches, sculptures, and community gardens (Saleebey, 2004); stairways in apartment complexes and common access paths to buildings (Festinger, Schachter, & Back, 1950); public gathering places (Semenza, 2003); semiprivate spaces, such as front gardens, verandas (Skjaeveland & Garling, 1997), and front porches (Brown, Burton, & Sweaney, 1998); sidewalks (Jacobs, 1961); local amenities (Altschuler, Somkin, & Adler, 2004); open spaces/parks (Cattell, Dines, Gesler, & Curtis, 2008); attractive places to walk; and local meeting places like shops and cafes (Baum & Palmer, 2002). It is the latter two, or the stores, cafes, coffee shops, and other “hangouts,” the “third places” of social interaction between one’s work and home life, to which Oldenburg and Brissett (1982) draw particular attention. These authors believe that it is the gradual disappearance of such “third places” that is linked to the larger changes in America’s built environment toward increased suburban sprawl and dependence on the automobile (Oldenburg & Brissett, 1982); in turn, suburban sprawl has led to the emergence of neighborhood physical environments that are “hostile to informal, public sociability” (Hummon, 1991, p. 932). Relatedly, Cattell and associates (2008) draw on evidence indicating that “everyday public [open] spaces” in neighborhoods are not only a resource and environment for social interaction but also possess significant benefits for residents’ enhanced well-being and quality of life at large (Cattell et al., 2008).
Built Environment and Walkability
Implicit in a number of studies presented above is the belief that community built environment that facilitates neighborhood sociability is one that is pedestrian friendly and replete with “third places” one could easily access on foot. A number of recent studies have examined the effects of built environment and neighborhood walkability. Here, a walkable neighborhood is defined as a community “full of pedestrian infrastructure and destinations” (Forsyth & Southworth, 2008, p. 2) such as streets and sidewalks pleasant for walking, destinations in close proximity (i.e., within walking distance) to residences, and safe for pedestrians. In particular, Southworth (2005) believes that there are six essential criteria for a walkable or pedestrian-friendly community: (a) varied land use, including a mix of commercial and residential buildings; (b) well connected to modes of public transportation; (c) walking paths of good quality (well lit, free from obstacles, etc.); (d) safety of the walking environment; (e) interesting and aesthetically appealing context or environment around the walking paths; and (f) well-connected network of walking paths, both within and beyond the community. Choi, Kim, Min, Lee, and Kim (2016) call the set of such characteristics “human-centered street design” (Choi et al., 2016, p. 120). Using a measure of pedestrian perceptions, these authors identified the following human-centered design elements as most significant for pedestrian satisfaction: streets with fewer vehicle lanes, presence of crosswalks, a small number of bus stops and driveways to obstruct pedestrian traffic, wider sidewalks, strategically placed planting strips, and commercial areas with easy access to pedestrians and people using public transit (Choi et al., 2016). Other studies, however, have not been so equivocal in their determination of “pedestrian-friendly” characteristics of streets and neighborhoods. For instance, residential density and street connectivity is associated with increased pedestrianism in some studies, but not in others (Lovasi et al., 2008). One of the reasons such inconsistencies may be the lack of clear differentiation between walking for exercise and walking for transportation, or recreational and utilitarian (or destination) walking, respectively (Kang, Moudon, Hurvitz, & Saelens, 2017). Utilitarian and recreational walking differ in the purpose, length, and speed of walking. Compared with recreational walking, utilitarian walking generally involves longer distances, higher speed, and has a clear destination for the walk; in general, utilitarian walking is more likely to occur outside the neighborhood of one’s residence (Kang et al., 2017). In comparison with recreational walking, the role of the built environment may be more significant for the utilitarian walking behavior. In the study by Kang and associates (2017), utilitarian walking tended to happen in non-home-neighborhoods, in communities with higher levels of residential and job density and less slope than one’s home neighborhoods. In contrast, for recreational walkers, the non-home-neighborhoods tended to be less residentially dense, have higher property values, and have larger area covered with parks and trails than their home neighborhoods.
All the environmental factors influencing walking behavior are not created equal, however. Alfonzo (2005) highlights the possibility of a “hierarchy of walking needs” when it comes to an individual’s decision to walk (Alfonzo, 2005, cited in Adkins, Dill, Luhr, & Neal, 2012). Adkins and associates (2012) cite empirical evidence about the primary importance of the distance and to one’s destination, and its pedestrian accessibility as factors influencing walking behavior; perceived safety, comfort, and aesthetic appeal of the walking environment are of secondary importance (Saelens & Handy, 2008).
Despite the inconsistency of empirical findings and the complexity of the relationship between built environment and pedestrian activity, a growing number of citizen groups have taken a stand against the automobile centeredness of the conventional norms, and advocate for walkable communities and “human-centered [urban] design” (Choi et al., 2016). Prominent among those groups is Congress for the New Urbanism, promoting the increasingly popular new urbanist urban planning paradigm.
New Urbanism and Traditional Neighborhood Design
New urbanism has an explicit goal of enhancing quality of life and building neighborhood social capital through transformation of community built environment. Inspired by the elements of traditional neighborhood design, new urbanists challenge the suburban sprawl and reliance on the automobile by promoting settings resembling the traditional American towns. The core characteristics of such settings are pedestrian-friendliness, traditional neighborhood structure with discernible edge and center, quality architecture, interconnected street grid networks, mixed-use housing, access to basic amenities within 5- to 10-min walk from home and work, and availability of housing at various price levels (Congress for the New Urbanism, n.d.). Interactions between the core dimensions of the new urbanist design are expected to encourage community diversity, and produce high levels of neighboring and bridging social capital among diverse neighborhood residents. In particular, the new urbanist ideals of high levels of neighborhood population density, a mix of residences and workplaces in a neighborhood, along with the availability of housing at variety of prices, would enhance community diversity, and increase in overall population flow and intensity. It is hypothesized that in a new urbanist neighborhood, walking and pedestrian street activity is encouraged by the design elements of mixed residential–commercial land use (which makes it more likely to see people on the streets during all hours of the day), local access to neighborhood facilities, and banishment of cars and garages to back alleys, along with enhanced public transit options, street grid that slows down traffic, and pleasant streetscapes comfortable for walking. In turn, pedestrian activity in the neighborhood (along with the increase in community population density and flow) will make it more likely that the residents will spontaneously “bump into” each other, leading to informal interaction and formation of community social capital (Brown & Cropper, 2001; Leyden, 2003). Here, the proponents of new urbanism rely largely on what has become to be known as social ecological thought (Schorr, 1963; Whyte, 1980).
New urbanist thought could also be linked to what is termed opportunity theory (Jacobs, 1961). Opportunity theory emphasizes the importance of community sociability and informal social control encouraged by broad sidewalks, strategic placement of parks, and mixed land use to encourage variety of functions, which, in turn, would create an environment conducive to activity on the street at all hours of the day, and consequently reduce opportunities for crime. Jacobs (1961) describes how neighborhood residents’ perception of community safety is enhanced by increased pedestrianism and continuous pedestrian use of public places, to give residents “eyes on the street” for informal social control. Specifically, Jacobs listed “four conditions for ‘vital urban life’” (Sung, Lee, & Cheon, 2015, p. 1): mixed residential–commercial land use to encourage pedestrian activity at all hours of the day; short street blocks, to slow down traffic and shorten walking distances for pedestrians; diversity of buildings (diverse in age and form) to provide opportunities for various types of jobs; and, along with the previous three, “a sufficient concentration of buildings” to encourage their use (Sung et al., 2015, p. 2). Beyond these “necessary conditions,” Jacobs also encouraged the creation of easily accessible neighborhood parks and public transportation.
To recap, new urbanists expect that traditional neighborhood design elements would contribute to the development of community social capital, both bridging capital of diverse ties between neighborhood groups and bonding capital of close neighborhood ties (Congress for the New Urbanism, n.d.). It may be hypothesized that social capital is developed by the spontaneous social interactions encouraged by the pedestrian-friendly, mixed residential–commercial neighborhoods, where all the essential community amenities (such as public transportation, neighborhood parks, and recreation opportunities) are available within a 5- to 10-min walk from one’s residence. Following Wood and Giles-Corti (2008), and contrary to Jacobs (1961), this author believes that community safety is more important as a context for increased social interaction rather than its outcome. Furthermore, the general appearance of the community appears as an essential condition for resident sociability and increased pedestrianism, along with easily accessible “third places” or spaces for neighborhood interaction. Beyond the objective features of one’s community, the author appreciates the notion that residents’ perceptions of their neighborhood environment (e.g., perceived safety and problems in the community, perception of access to local amenities and services, etc.) may be a key element in their sociability and the development of neighborhood social capital at large.
Social Capital, New Urbanism, and Traditional Neighborhood Design
A number of empirical studies have examined social capital in traditional neighborhood developments and new urbanist communities; however, the evidence to what extent such urban form has an effect on residents’ social capital is mixed. Leyden (2003) found that in Galway, Ireland, a traditional neighborhood, community walkability was linked to higher levels of social capital among its residents. As a caveat, however, Lund (2003) concluded that besides the pedestrian friendliness of the neighborhood, neighbor interactions (one of the forms of social capital) were also predicted by residents’ perceptions about the importance or value of neighboring. The results of the Podobnik’s (2002) study of the Orenco Station of Portland, Oregon, a traditional neighborhood development, revealed that Orenco Station had higher level of residents’ bonding social capital (as trust in neighbors) than the comparison community, but lower level of bridging social capital (measured as residents’ attitudes toward ethnic and class diversity). Somewhat differently, Hampton (2002, cited in Sander, 2002) used a lost-letter experiment to gauge the level of social capital in the 71 communities studied. In each of these communities, 60 lost letters were distributed; in the end, the new urbanist communities of Celebration and Seaside, Florida, led the list of “lost” letters returned unopened (Sander, 2002). Conversely, another author found support to the new urbanist thesis that suburban sprawl “suppresses” neighborhood social capital. In his study, longer commuting time led to decrease in trust in neighbors, and residents of communities with greater proportion of solo commuters were less likely to participate in civic activities (Williamson, 2002, cited in Sander, 2002). Interestingly, the results of the research by Wood and associates (2008) contradicted the findings of the earlier studies: In Perth, Australia, neighborhoods with a new urbanist street grid had lower levels of social capital than conventional suburbs (Wood et al., 2008).
The empirical studies reviewed above use a variety of measures for measuring the concepts of social capital and built environment, which may be one reason for their contradictory findings. For social capital, generally attitudinal measures are used; urban design elements may be operationalized by objective or perceptual indicators. To contribute to the research on the topic, the purpose of the present study was to examine the role of residents’ perceptions of the aspects of built environment in social capital development by comparing two Chicago, IL, area communities: Bucktown, a community characterized by principles of traditional neighborhood design, and the Village of Schaumburg, a neighborhood that could be considered an example of a large and well-established suburb. Furthermore, this author sought to develop suggestions for future research about the residents’ perceptions of the built environment as determinants in community differences in social capital, with a focus on perceptions of safety and community problems, neighborhood walkability, and local access to basic amenities.
Method
In this study, the dependent variable was social capital; independent variables tapping into residents’ perceptions of the built environment and community safety were (a) perceived access to basic amenities within neighborhood boundaries, (b) perceived pedestrian friendliness of the community, (c) perceived neighborhood crime, and (d) perceived community problems. The demographic characteristics of the sample served as independent or control variables in this study. Based on the relevant literature, the following demographic variables were included in data analysis: gender, age, income, education, home ownership, years lived and years expected to live in the neighborhood, political orientation, religious affiliation, and frequency of religious service attendance.
Data Collection
This research was based on two data sources, hereafter referred to as Study A and Study B. Study A, a research by VanderWaal (unpublished), was used as a source of secondary data. The data for Study A were collected using a cross-sectional 10- to 15-min telephone survey of a random sample of 709 adults residing in two Chicago area neighborhoods, Bucktown and Schaumburg, between February and August of 2005. The researchers for Study A had selected Bucktown as a community exemplifying traditional neighborhood design principles, and the Village of Schaumburg as an example of suburban sprawl. The boundaries for the two neighborhoods were based on census tracts, and subjectively determined by the principal investigator (PI) of Study A. Following the data collection and analysis for Study A, the author, or PI for Study B called between December of 2005 and May of 2006 the 485 individuals who, at the end of the survey for Study A, had agreed to be contacted for a longer follow-up phone interview. In total, 197 respondents, 97 from Bucktown and 100 from Schaumburg, completed the Study B 30- to 45-min semistructured interview. Note that the Study A data were used only for these 197 respondents, and not for the entire Study A sample of 709 (their phone numbers were used as identifying information). Thus, the final sample size for the study described below was 197, including 97 participants from Bucktown and 100 from Schaumburg.
Bucktown, a part of the historic Wicker Park District of Chicago dating back to 1837 (Coorens, n.d.), is a densely populated, pedestrian-friendly Chicago neighborhood with mixed commercial–residential land use, and easy access to public transportation and amenities such as restaurants, stores, and schools within 5- to 10-min walk from one’s residence. A neighborhood northwest of the Loop district (Chicago Traveler, 2017), Bucktown encompasses 1.021 square miles, has a population of 24,347 and population density of 23,836 people per square mile (City-Data.com, 2017). For more information on the demographic profile of Bucktown and Schaumburg, see Table 1 below. Note that since the U.S. Census Bureau (n.d.) has only limited data available at the census tract level, zip codes that most closely approximated the neighborhood boundaries were used for retrieving the demographic information: 60,647 for Bucktown and 60,173 for Schaumburg.
Bucktown and Schaumburg Demographic Characteristics, U.S. Census 2010.
The Village of Schaumburg, incorporated in 1956 (Village of Schaumburg, n.d.), is a Chicago suburb of 19.2 square miles, with an estimated population of 74,446 in 2016 and much lower population density of 3,862 (U.S. Census Bureau, n.d.). As one of the largest suburbs of Chicago, Schaumburg is known for its landmarks: The gigantic Woodfield Mall, as well as Ned Brown Preserve and nearby Legoland Discovery Center (google.com, 2017).
Measures
Neighborhood social capital
Twenty-two indicators from Study A survey tool, and 31 from Study B questionnaire, measured neighborhood social capital. For further analysis, the author combined these indicators in eight conceptually identifiable factors by using a principal components (factor) analysis procedure. From the items included, the varimax rotation extracted 12 components. The seven components that individually captured 4.5% or more of the variance were included in further data analysis as measures of social capital; an eighth factor that captured 3.82% of the variance was also added as it was conceptually significant. Together, the eight factors captured 49.65% of variance in the principal components analysis and are listed below. See Table 2 for the factor loadings (in bold) of the survey items associated with each social capital factor.
Knowing neighbors and informal neighboring: The first factor that emerged from factor analysis, knowing neighbors and informal neighboring, had an eigenvalue of 8.74 and accounted for 9.48% of the overall variance. This factor included indicators about respondents’ reports of recognizing and knowing people in one’s neighborhood, and casual neighbor interactions.
Perceived neighbor support: The variables on this attitudinal social capital factor tapped into respondents’ reported perceptions that one’s neighbors are supportive, dependable, and willing to take collective action to solve neighborhood problems.
Informal problem solving: This dimension of social capital was based on participants’ answers to survey questions talking informally to other neighbors about neighborhood problems, interaction with one’s nearest neighbors, and informal assistance given and received in the neighborhood.
Close ties and social divisions in the neighborhood: This factor included the elements of the self-reported numbers of respondents’ closest friends in the neighborhood and frequency of close neighbor interactions. Interestingly, the variable of perceived social tensions in the neighborhood loaded positively on this social capital factor as well, suggesting perhaps that having one’s closest friends reside in the neighborhood may also lead to the formation of “cliques” or pockets of close networks at odds with each other.
Group participation: Respondents’ reports of their participation in formal and informal extra- and intra-neighborhood groupings, including formal organizations and more informal book clubs, hobby groups, neighborhood associations, and others.
Mingling between neighborhood groups: Perception of opportunities for community participation, and the level of informal interaction between diverse neighborhood groups.
Perceived parental relationships: Linked to the Coleman’s (1988) concept of closure in parental networks, this factor captured the indicators linked to one’s perception of the extent that parents in the community know each other and their children’s friends.
Sense of collective power and social obligation: Similar to Bandura’s (1997) definition of self-efficacy, this factor assessed the respondents’ perception of their capacity to affect neighborhood decisions through collective action. The component was also associated with one’s acceptance of the construction of affordable housing in their community.
Factor Analysis Results: Factor Loadings of Social Capital Factors.
Note. The items from the Study A survey are displayed in italics; the nonitalicized items are from the Study B survey questionnaire.
Demographic variables
The data analysis sought to control for demographic differences between Bucktown and Schaumburg neighborhoods. To determine the differences, the following demographic variables were used: age, gender, education, income, home ownership, number of children living at home, one’s living arrangement (a single person or couple, with or without children), political orientation, religious affiliation, frequency of religious service attendance, length of residence in one’s neighborhood (in months), and the length of time one planned to continue living in their current residence.
Residents’ perceptions of the built environment
Two variables—perceived availability of basic amenities and services within the boundaries of the neighborhood, and residents’ perceptions of neighborhood walkability or pedestrian friendliness—were used as predictors in multiple regression analyses, to explore the relationship between perceived built environment and neighborhood social capital. Perceived access to basic amenities was measured by the unstandardized Local Amenities Scale; the Destination Walking Scale (DWS) served as a proxy measure of the perceived pedestrian friendliness of the community. For both scales, each respondent received an additive index score by summing their responses for the scale items.
Perceived access to local amenities
Adapted from the Community Survey Questionnaire of the Project on Human Development in Chicago Neighborhoods, the seven-item Local Amenities Scale (Cronbach’s alpha = .44) asked the respondents questions about the availability (yes or no) of the following amenities in their neighborhood: an open space, community newsletter, a crime prevention program, family health service, mental health center, banking services, and a group dealing with local issues. Each “yes” was coded as “1,” and “no” as “0.” For each participant, their total summed score on the scale could range from a low of 0 (the respondent reporting no community amenities available) to a high of 7 (the respondent reporting all seven neighborhood amenities on the survey as available).
Perceived neighborhood walkability
Respondents’ scores on the DWS modeled after the E-Neighbors Survey (Massachusetts Institute of Technology, Department of Urban Studies and Planning, n.d.) were used as a proxy indicator of neighborhood walkability or pedestrian friendliness. The DWS (Cronbach’s alpha = .78) asked the respondents whether they had walked (yes—coded as “1,” or no—coded as “0”) to six different community destinations in the previous month: a friend’s home; a retail (not a grocery) store; a park or a place for recreation; a restaurant, movie theater, or a club; a public transit stop; and/or around the neighborhood for relaxation. For each participant, their total score on the scale could range from a low of 0 (the respondent reported not having walked to any of the destinations on the survey during the previous month) to a high of 6 (the participant reported having walked to all six destinations within the previous month).
Perceived crime and community problems
As described previously, a number of authors have linked neighborhood social capital to two perceptual factors, including perceived safety (or crime) and community disorder (Wood & Giles-Corti, 2008; Wood et al., 2008). The residents of a neighborhood perceived as safe, with few (or no) signs of disorder (e.g., litter, graffiti, loitering, etc.), seem to be more likely to “venture out” of their home and interact with their neighbors. Hence, besides the indicators of residents’ perceptions of the built environment described previously, the exploratory regression analyses included measures of residents’ perception of community safety and problems as independent predictors of social capital. Two nonstandardized scales were used to measure perceptions of crime and community problems, the Perceived Crime Scale and Community Problems Scale. For every respondent, two index scores were created by summing their responses on each of these scales.
Perceived crime
Following the study by Perkins and Taylor (1996), the five-item Perceived Crime Scale (Cronbach’s alpha = .85) tapped into the frequency (often—coded as “3,” sometimes—as “2,” rarely—as “1,” or never—as “0”) the respondents had witnessed five types of crime or disorder in their neighborhood in the previous 6 months: A fight where a weapon was used, a violent argument between neighbors, a gang fight, a sexual assault, and/or a robbery or mugging.
Perceived community problems
As a proxy indicator of neighborhood problems, participants’ perception of the seriousness of neighborhood problems, the 10-item Community Problems Scale (Cronbach’s alpha = .66) adapted from the General Household Survey asked the respondents about a variety of potential community problems: litter on the streets, graffiti, vacant buildings, loitering, burglaries and other crimes, drug sales the respondent had witnessed, divisions between neighborhood groups, neighbors who “cause trouble,” “wrong kinds of people” moving in, and lack of public transportation. For each item, it was asked, “How much of a problem is . . . ?” The response options included “a big problem” (coded as “2”), “somewhat of a problem” (coded as “1”), and “not a problem” (coded as “0”).
Data Analysis
The data were analyzed by using SPSS, version 15. Crosstabs and a series of chi-square tests were used to describe the demographic characteristics of Bucktown and Schaumburg, and determine any statistically significant demographic differences between the two neighborhood samples. After identifying the key social capital factors by using the principal components (factor) analysis, the author conducted a series of independent-samples t tests to compare Bucktown and Schaumburg on their sample means for these factors, followed by a two-step multivariate regression analysis to examine whether the differences in the means (if any) would be maintained after controlling for demographic differences between Bucktown and Schaumburg samples. Finally, the researcher used a (one-step) multivariate regression analysis to determine the indicators of residents’ perceptions of the built environment that emerged as statistically significant predictors of social capital factors.
Results
Demographic Characteristics of Respondents
The study sample included 100 respondents from Schaumburg and 97 from the Bucktown neighborhood. As shown in Table 3, the participants from Bucktown and Schaumburg were similar in a number of demographic characteristics, including ethnicity, gender, number of children below 18 years in the household, living arrangement, and number of years lived in the neighborhood. Table 3 also shows that there were some notable demographic differences between the participants from Bucktown and Schaumburg, in variables that were controlled for in further data analysis. The results of the chi-square tests indicated that the differences between the Bucktown and Schaumburg samples were statistically significant for the variables of age, education, annual household income, home ownership, the length of time one expected to live in their residence, political orientation, religious affiliation, and frequency of religious service attendance. The respondents from Bucktown were notably younger, more educated and affluent, and more likely to identify as nonreligious. Participants from Schaumburg were more likely to own their home, plan to stay in their residence for 4 years or more, identify as moderate, and attend religious service at least once a month.
Demographic Profile of the Study Participants From Bucktown and Schaumburg.
p < .05. **p < .01, two-tailed.
Differences in Social Capital
To evaluate the differences between the levels of social capital among Bucktown and Schaumburg respondents, a series of independent-samples t tests were used to compare the means in the factor scores for the eight social capital factors listed above. As indicated in Table 4, of the eight major social capital factors, t-test comparisons of mean scores of the two communities were statistically significant for three: perceived neighbor support, t(195) = −3.54, p < .01; informal problem solving and mutual aid between neighbors, t(195) = 3.74, p < .01; and close ties in the neighborhood (which also included the perception of social divisions), t(197) = 3.37, p < .01. Respondents from Bucktown reported higher levels of two forms of neighborhood social capital: informal problem solving, and close ties between neighbors. In contrast, Schaumburg respondents had a higher level of one form of social capital, perceived neighbor support. These differences were maintained after controlling for the demographic differences between the neighborhoods in a follow-up two-step hierarchical multiple regression analysis, whereby the demographic characteristics that had been significantly different between neighborhoods were inserted in the first step, followed by the variable of neighborhood type—traditional or suburban—in the second step. (The results of the regression analysis are available from the author upon request.)
The t Test Comparisons of Social Capital Factors: Bucktown vs. Schaumburg.
p < .05. **p < .01, two-tailed.
Correlates of Neighborhood Social Capital
What other findings about the relationship between social capital, perceived crime and problems in the community, and neighborhood residents’ perceptions of the built environment do the data suggest? To answer this question, the author conducted three exploratory multiple regression analyses on the social capital factors that were significantly different for Bucktown and Schaumburg participants: perceived neighbor support, informal neighborhood problem solving, and close neighborhood ties.
Predictor variables in regression models
As the literature does not generally differentiate between the determinants of different forms of social capital, the researcher used an identical set of predictor variables in the three regression models.
Demographic predictors
As some of the demographic variables commonly cited as predictors of social capital were already included in the two-step regression analyses controlling for demographic differences between neighborhoods (described above), only the individual demographic characteristics of gender, length of time lived in the neighborhood, and having children living in the household were used as demographic predictors. Ethnicity was excluded from the analysis due to the low level of ethnic variability in the sample. As all the demographic predictors were categorical, they were recoded into dummy variables for the analysis, whereby each “yes” was coded as “1,” and “no” as “0.”
Perceived built environment variables
To explore the perceived environmental predictors of social capital, the regression analyses also included respondents’ scores on the Access to Amenities and Destination Walking Scale as independent predictors, along with their scores on Perceived Crime and Neighborhood Problems Scale. As shown in Table 5, the results of the independent-samples t-tests conducted prior to regression analyses indicated that the differences between the respondents from Bucktown and Schaumburg were statistically significant for all four scales. Bucktown participants scored higher in perceived access to amenities and neighborhood walkability (destination walking) than Schaumburg residents. They were also likely to report more community problems, and give a higher score for the frequency of crime witnessed in the previous 6 months. Neighborhood differences in the perceived frequency of crime are consistent with online statistics. According to Niche.com (2017a), in Bucktown, for every 100,000 residents, there were 267 assaults, zero murders, 236 robberies, 771 burglaries, 4,206 thefts, and 330 motor vehicle thefts in 2016. In contrast, in Schaumburg, there were 31 assaults, one murder, 24 robberies, 169 burglaries, 2,201 thefts, and 55 motor vehicle thefts per 100,000 residents in 2016 (Niche.com, 2017b).
The t Test Comparisons of Hypothesized Environmental Predictors: Bucktown Versus Schaumburg.
p < .05. **p < .01, two-tailed.
The results of multiple regression analyses
The results of the regression analyses are presented in Table 6. Each built environment predictor in the regression models emerged as statistically significant predictor of one of the three social capital factors in the analyses (except for the variable of perceived crime that predicted two). For four perceived built environment predictors, (a) perceived neighborhood crime was
Regression Analyses of the Hypothesized Predictors of Social Capital.
p < .05. **p < .01, two-tailed.
In the regression analyses, only two demographic variables emerged as statistically significant predictors of one or more social capital factors. (a) Presence of children in one’s household was
Based on the results of the exploratory multiple regression analyses, the following hypotheses may be proposed for future research: Perception of one’s neighbors as supportive increases in the context of less perceived neighborhood problems, lower level of perceived safety, and improved access to basic amenities in the neighborhood. Furthermore, perceived neighborhood walkability appears to be a predictor of some forms of neighborhood social capital but not others—in this study, it was correlated with respondents’ informal problem solving and close neighborhood ties. It may be that the more residents are likely to walk in the neighborhood, the greater their likelihood of engaging in “informal problem solving”—talking to other neighbors about community issues and offering help or receiving assistance from their neighbors (note that this correlation was only marginally significant in the regression results). Following new urbanist expectation, it could be similarly hypothesized that increased pedestrianism in the community—if it is coupled with an increase in population density—will promote the formation of close social ties between residents in the community, due to the increased informal interactions from neighbors. In contrast, having children living in the household may be associated with residents having less close ties in the community but may be an important motivating factor for neighbors to engage in mutual aid interactions of informal problem solving.
Discussion
The findings of this study demonstrated that the participants from Bucktown, the traditional neighborhood in this study, reported higher levels of social capital in two areas: They were more likely to engage in informal problem solving and have more close relationships with other neighborhood residents. However, the results also indicated that in the suburban neighborhood, respondents saw their neighbors as more supportive than the participants from Bucktown. What could account for these results?
Perceived Neighbor Support
A number of reasons could be suggested for the higher level of perceived neighbor support found in Schaumburg. Importantly, a closer look at the survey items that had a factor loading of 0.5 or higher on the social capital factor of perceived neighbor support (see Table 2 for more detail) reveals their similarity to some traditional indicators of community cohesion and informal social control—feeling close to one’s neighbors and the need to be respected in the neighborhood, and certainty that one’s neighbors would “keep their eyes open for possible trouble” and get together “if there was a [community] problem.” One commonly cited correlate of perceived community cohesion is residential stability. Although there were no statistically significant differences in residential stability between the Bucktown and Schaumburg samples, there was a significantly larger portion of participants from Bucktown who expected to continue living in the same neighborhood for no longer than 1 to 3 years; in turn, in comparison with Bucktown, a larger share of the Schaumburg sample indicated that they anticipated remaining in their neighborhood for at least another 4 years or longer. However, this variable (“expected” residential stability) was not a significant predictor of perceived neighbor support in the multiple regression analysis. It may be surmised that beyond differences in “expected” residential stability, Bucktown, as a neighborhood in downtown Chicago, had a larger number of nonresidents entering the neighborhood (for business, commercial, or other reasons) on a regular basis. In contrast, it could be expected that as a suburb, land use in Schaumburg is more compartmentalized between commercial and residential areas, and therefore, less outsiders would be seen in the residential parts of the community. Would regular contact with nonresidents be likely to undermine community cohesion? This is a question that was not studied in the current article, but raises an interesting hypothesis for future research.
The higher social capital in the form of (perceived) supportiveness of one’s neighbors in the suburban community of the study could also be explained by the notably lower level of perceived crime and less perceived community problems found among participants from Schaumburg when compared with those from Bucktown. In the regression analysis, perceived community problems and perceived crime were negatively associated with the variable of perceived neighbor supportiveness. This finding supports the results of research demonstrating the positive association between perceived neighborhood safety and trust in one’s neighbors (Perkins, Florin, Rich, Wandersman, & Chavis, 1990; Ziersch, Baum, Macdougall, & Putland, 2005). It may also suggest, as some researchers have pointed out (Wood & Giles-Corti, 2008; Wood et al., 2008), that viewing one’s neighborhood as safe, clean, and quiet, with few or no signs of disorder, may be important in the process of community social capital formation. Safe environments may make it more likely that residents venture out and interact with their neighbors, creating an environment that would allow social capital to flourish (Wood et al., 2008). Conversely, viewing one’s neighbors as supportive, or believing that they are likely to “watch out” for other residents, may contribute to the perception of neighborhood environment as safe. Hence, perception of neighborhood safety is often seen as an outcome of increased social capital and informal social control in the community (Sampson et al., 1997). Due to the cross-sectional nature of this study, the causality in the association between perceived crime or safety, and perception of the supportiveness and dependability of one’s neighbors cannot be determined—in fact, as some authors suggest, perhaps a reciprocal or two-way relationship between perceived safety and social capital is likely (Wood et al., 2008).
An important exploratory finding of this study showed that the respondents’ perception of the availability of the basic amenities (a health service, mental health center, banking service, a park or recreation area, neighborhood organization or watch, and a community newsletter or newspaper) in one’s neighborhood was positively correlated with the perceived supportiveness of neighbors. This association may be easy to explain, as using neighborhood facilities may provide positive contexts where some forms of social capital can be created or developed. Visiting a local library, a health center, or a neighborhood bank, or participating in a neighborhood association may increase opportunities for neighbor interaction, and help to improve the relationships between neighbors and perhaps increase the perception of neighbors as dependable and supportive. Several previous studies have identified access to local services as a factor essential in the development of residents’ social capital, including perceived trust, norms of reciprocity, and mutual aid between neighbors (Baum & Palmer, 2002; Cattell, 2001).
More generally, the higher level of perceived neighbor support found in Schaumburg (and lower level of two other forms of social capital, informal problem solving and close neighborhood ties) highlights the importance of defining and measuring neighborhood social capital as a multidimensional concept. In this context, the “norms” dimension of social capital (e.g., perception of neighbor supportiveness) may be equally critical to its perceived structural dimensions of community networks and participation—particularly an important consideration in many American neighborhoods where the actual forms of interaction and friendship between neighbors may be declining.
Informal Problem Solving
Although the respondents from the traditional neighborhood in this study saw their neighbors as somewhat less supportive than the residents of the suburban Schaumburg, they were more likely than the participants from the suburban community to report talking to their neighbors about neighborhood problems, and engaging in other informal problem solving and mutual aid interactions. The higher levels of perceived crime found among Bucktown respondents could be partly responsible for more informal problem solving reported in this neighborhood, as community crime issues may galvanize its residents to take collective action and offer mutual aid.
Although less helpful for explaining neighborhood differences in social capital, it is important to observe that in this study having young children in the household was a significant (positive) predictor of informal problem solving. This is an important finding that highlights the potential impact of family-friendly community policies in neighborhood social capital development. Even though research findings on the role of parental relationships in community social capital generation are not unequivocal (Putnam, 2000), studies have observed that young parents are generally more involved in neighbor interactions, as children tend to provide a platform for neighbor support and reciprocity (Lund, 2003; Wood et al., 2008). From another angle, the general family-friendliness of the neighborhood is likely to attract young families and lead to higher levels of community social capital among its residents, as parents develop relationships and engage in mutual aid interactions. More generally, the positive association found in this study between informal problem solving and presence of young children in the household highlights the role of children and parental relationships as a potentially significant source of neighborhood social capital.
Close Ties and Perceived Divisions in the Neighborhood
In this study, Bucktown participants had higher levels of the form of social capital that tapped into close ties and interactions between neighbors. As Bucktown has also significantly higher population density than Schaumburg, and a greater mix of businesses and private residences in the neighborhood, a simple increase in human interactions due to greater population flow and intensity in this neighborhood may have contributed to this finding. Furthermore, as Bucktown residents perceived their community as more walkable or pedestrian-friendly community, higher level of close ties between neighbors and an increase in resident interactions may be linked to the positive correlation found between perceived walkability and having close ties in the neighborhood (as described above). It may be surmised that a higher level of reported pedestrian activity in Bucktown, and simply more people being present at all hours of the weekdays, led to a more frequent contact between the participants from this community, which may have resulted in an increase in their social interactions and formation of close ties.
In contrast to the positive correlation found between the variables of walkability and close ties/neighbor interactions, having young children in the household was a negative predictor of the social capital factor of close ties. It may be that parents of young children are more likely to engage in casual neighbor interactions and more superfluous neighbor reciprocity, and forming one’s close ties with nonresidents rather than the people living in the neighborhood. Albeit contradictory, both findings have important implications for social capital theory. Further studies are clearly needed to examine the relationship between the variables of having young children in the household and close ties among neighbors, and the role of community walkability in the process.
Interestingly, in factor analysis, residents’ perception of social divisions or (nonethnic) tensions in the neighborhood was also a component of the social capital factor of close neighborhood ties. It may be hypothesized that having one’s closest friends reside in the neighborhood and engaging in close neighbor interactions may lead to pockets of dense neighbor networks in the community. Besides the positive implications of close neighborhood ties, it may also result in the formation of some “cliques,” dividing the broader community and undermining its cohesion.
To recapitulate, in this study, some aspects of the respondents’ perception of the built environment and community context (i.e., perceived safety and community walkability) were positively related to community social capital (note that the effect of the environmental variables was fairly small). Further studies are needed to explore the pathways whereby the perceived built environment may contribute to social capital formation, with an ultimate goal of developing a comprehensive model of the predictors and correlates of neighborhood social capital.
Limitations of the Study
This study had a number of limitations. The sample was relatively small, as it was limited to the 197 individuals out of original 709 who had completed the survey for Study A. Furthermore, as the follow-up interviews lasted 30 to 60 min (or more), extra time and effort was required from the respondents to complete the interview. Therefore, willingness to finish the phone survey may have introduced a selection bias in the study sample, by including a potentially disproportionate share of individuals in the sample who may have been different in some characteristics from the general neighborhood population (i.e., being more motivated and committed to the neighborhood, with possibly higher levels of social capital). Relatedly, nonresponse bias may have been added due to the time difference (ranging from 4 to 16 months) between the data collection for Study A and Study B. The author was not able to reach 82 of the 485 respondents from Study A of this study who had agreed to participate in a follow-up interview. In parallel, the time difference between the data collection for Study A and Study B poses the question of reliability as respondents’ answers could have changed over the course of this time. A longitudinal study tracking the same sample over time and asking them identical questions at different points of time may be recommended for a future study seeking to enhance the reliability of the findings.
Due to the cross-sectional nature of the study, causality between predictors and the hypothesized outcomes of social capital could not be established. For instance, perceived safety can be seen as a factor necessary for community social capital formation, but could also be viewed as its outcome. Once again, a longitudinal study would be needed to substantiate any reciprocal relationships, time order and causality.
Measurement is a key limitation of this study, particularly the reliability and validity of its survey tool. The questionnaires for Studies A and B included items from a variety of survey instruments in an attempt to develop a comprehensive measure of neighborhood social capital. Drawn from several sources, these questions could have originally been developed for different populations and geographic locations. Furthermore, the perceptual indicators for social capital used individual-level data to understand community-level social capital in this study. Following Lin (2001), it is this author’s belief that using measures of neighborhood social capital that examine the properties of residents’ networks could be useful in future studies, particularly those based on more complex network analysis and multilevel modeling methodologies. Furthermore, utilizing a larger number of variables to study perceived built environment, along with more nuanced measures of this concept (including, among others, additional perceptual indicators of neighborhood pedestrian friendliness and more objective observational measures of walkability), could be useful for a future study seeking an in-depth understanding of the relationship between residents’ perceptions of the built environment and neighborhood social capital.
Similar to the social capital and built environment variables, access to amenities and community problems and safety were measured with perceptual indicators, which may have resulted in a bias in the data collected. When asked about their behaviors and attitudes, people might answer in a way that is not fully consistent with their actual practice as one tends to answer as they wish they would have behaved. Other, more objective measurement methods could be used to capture these variables—particularly the availability and use of community amenities. The author suggests that more objective indicators could be suggested for the measurement of community problems and amenities in future studies, particularly observational measures and existing statistics.
The most notable limitation of the validity of study findings is associated with the limited number of neighborhoods included in the study. It is difficult to make any generalizations based on the responses of participants from two neighborhoods. It is also essential to point out the marked changes in population composition in the process of regentrification that one of the study neighborhoods, Bucktown, was undergoing at the time of the data collection. Therefore, the Bucktown sample may have offered a temporarily inaccurate representation of the community cohesion and social capital held by the neighborhood residents. The community transformation in Bucktown also raises the broader question of the validity of using this neighborhood as a prototype of a traditional neighborhood. Architecturally, Bucktown is losing some of the characteristics of traditional neighborhood design, including the diversity of its population. Careful selection of a similar but more stable neighborhood is recommended for future research, along with an inclusion of a larger number of study neighborhoods for the comparison.
Conclusion
This study provides a window into the life of the residents of two Chicago area communities, Bucktown and Schaumburg. The results of the study help to highlight the promise of interventions involving a change in perceived built environment for facilitating the formation of neighborhood social capital. Yet the study cautions against concluding that there is a deterministic relationship between residents’ perceptions of the physical (built) environment and the social goals of a community. Hence, a comprehensive understanding of the predictors of neighborhood interactions and social capital networks is required for an effective analysis of the relationship between neighborhood social capital and perceived built environment—and perceived physical environment at large.
Footnotes
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.