«AmericAn neighborhoods: inclusion And exclusion Volume 16, Number 3 • 2014 U.S. Department of Housing and Urban Development | Office of Policy ...»
of affordable housing in mixed-income developments. Chaskin and Joseph (2011) reported that renters of affordable housing units are similar in life circumstances to former public housing residents, whereas owners of affordable units are similar to market-rate owners and renters. Tach (2009) focused on the differences between “newcomers” and “long-term” residents; although most long-term residents were in the lowest income stratum, both groups included some households with incomes that would qualify them for affordable housing. These studies suggest that former public housing residents and affordable-housing renters may face similar challenges to forming meaningful relationships with more affluent neighbors, but existing data are too limited to understand fully the particular opportunities or constraints that affordable-housing residents face. To understand the potential effect of these programs on the well-being of recipients and their potential utility as higher income neighbors in mixed-income settings, it is essential to develop a better understanding of their personal networks—their composition and range, the prevalence of ties to neighbors near and far, and the use value of these relationships for both getting by and getting ahead.
Network Structure Access to social resources is determined in part by the composition of one’s social network and the properties of the network as a whole, including the range of the network, strength of ties, level of reciprocity, and density of the network (Lin, 2000). The structure of networks facilitates some opportunities and behaviors and constrains others. Dense networks (wherein most of the individuals know one another and few others outside the group) are generally composed of similar individuals and characterized by high levels of trust and mutual obligation that foster the sharing of available resources and effective social control (Briggs, 1998; Coleman, 1988). The kind of interdependence that is typical of dense, bonding networks, however, can produce negative consequences for its members. Individuals can be overburdened by the demands of their obligations to others even when favors are likely to be returned, particularly in a setting where individuals are frequently in need of support because of precarious finances or personal instability. Curley (2009) reported that relocated public housing residents are less likely to form relationships with new neighbors to preserve precious resources and avoid potentially “draining” ties. Solidarity among group members who bond over shared adversity may face a downward leveling of norms, whereby individual successes are viewed as unlikely or impossible (Portes, 1998). Perhaps most importantly, dense networks are likely to convey redundant information and lack bridges to outside resources (Burt, 1992) such that advice and assistance lead to the reproduction, rather than the improvement, of life circumstances (Granovetter, 1995).
By contrast, wide-ranging networks comprising weak ties are more likely to serve individuals by broadening knowledge and access to information, facilitating connections to other resources through brokered ties, and generally increasing one’s competitive edge (Burt, 2001; Granovetter, 1973).
Strong and weak ties serve individuals and families in different ways, but it has been suggested that the presence of these bridging ties is particularly critical for low-income residents’ upward mobility (Briggs, 1998). Although weak ties have been identified as beneficial for securing work and job advancement (Granovetter, 1973), these types of relationships are less likely to provide sustained support to individuals, who thus may require a larger network to achieve the same levels of engagement provided by fewer strong ties. Larger networks require maintenance and may be less likely to fulfill obligations, making them costly for individuals to acquire and sustain (Burt, 1992).
50 American Neighborhoods: Inclusion and Exclusion Building Ties: The Social Networks of Affordable-Housing Residents Relationship Content Whereas network structure defines the extent of available resources (opportunity) and propensity for certain relations to be engaged, relationship content focuses on activated ties and the prevalence of use for specific instrumental action (Hurlbert, Beggs, and Haines, 2001). In this article, we examine three types of support, each of which may benefit low-income residents in different ways.
Emotional, or expressive, support includes those actions related to general caring, empathy, or sharing between trusted individuals or confidants. The presence or absence of this type of support has been shown to have both direct and indirect effects on well-being (Berkman, 1995; House, 1981) and may be particularly salient for helping low-income households cope with both acute and chronic stress (Thoits, 2011). Instrumental support is the provision of practical assistance, either in the form of small favors of more substantial commitment of resources. This form of assistance may convey critical resources to low-income residents who lack financial resources and frequently live at the margin, enabling individuals to acquire services or goods not otherwise attainable because of limited means (Edin and Lein, 1997; Venkatesh, 2006). Informational support is the provision of knowledge or information that enables people to help themselves.6 Obtaining knowledge through one’s network may be less costly than acquiring it on one’s own (Coleman, 1988); however, for lowincome households, the value of this type of support is likely contingent on whether it provides new information not otherwise available (Hurlbert, Beggs, and Haines, 2001) and on the extent to which it affords opportunities or advantages (Granovetter, 1995; Henley, Danziger, and Offer, 2005).
Affordable Housing and Social Networks Lower income households have been shown to have small, locally based networks that are primary sources of emotional and instrumental support (Campbell and Lee, 1992; Fischer, 1982; Stack, 1974). Residential mobility may disrupt existing neighbor networks, leading some households to make secondary moves to be closer to family and friends who provide emotional and instrumental support (Boyd, 2008). Research on relocated public housing residents’ exchanges with new neighbors is mixed. Some research shows that interactions in the new location are mostly casual and limited to exchanges within income and tenure groups (Chaskin, 2013; Chaskin and Joseph, 2011). Rasinski, Lee, and Haggerty (2010), however, showed that residents engage with new neighbors in a variety of activities related to help and advice, and most of the long-term residents studied by Tach (2009) reported instrumental support exchanges with neighborhood-based networks. Kleit (2010) found substantially lower rates of neighboring after relocation off site but reported little change in access to social support among English speakers. This finding underscores Haines et al.’s (2011) point that neighborhood ties make up a minimal proportion of the typical network and therefore should not be viewed in isolation from the broader set of social relationships and resources available.
Affordable-housing residents may be less socially isolated than the lowest income households that qualify for public housing or vouchers. Because income and social network size generally have a positive association, these less poor households may have larger social networks overall. Higher rates of labor force participation may provide opportunities for a wider range of relationships, In this article, informational support includes appraisal support, sometimes defined separately as the sharing of information that helps people evaluate themselves. For a discussion, see Tardy (1985).
including coworkers and employers. These same factors may also make it less likely that affordablehousing residents’ networks are locally bound—many or even most of their relationships may be with individuals who live in other parts of the city or country. If so, moving to affordable housing may not alter their relationships in any significant way. On the other hand, residents who share the experience of applying for affordable housing, move to a newly constructed building (and sometimes also a new neighborhood) within a few months of one another, and live in close proximity under the same roof may share enough common experiences to form relationships with one another.
Establishing relationships with neighbors may benefit affordable-housing residents even if they do not exhibit the kind of social isolation often associated with the most disadvantaged households. Forming local ties may generally increase residents’ sense of belonging and ease the transition to life in a new building and, in many cases, a new neighborhood. Ties to other low-income working neighbors who face similar challenges may facilitate the sharing of strategies and resources that help individuals and families to buffer stress and manage everyday challenges. Weak ties to neighbors, particularly with those who are better off, may augment existing relationships and thereby provide access to additional resources or new information that creates opportunities and promotes upward mobility over time.
Data and Methods Data were gathered from 120 residents who moved to a newly constructed affordable rental housing complex developed as part of NHMP. Study participants applied to a housing lottery that allocated 241 affordable rental units7 in two midrise buildings. Each of the 241 households that received housing through the lottery was recruited for an interview approximately 4 years after applying for housing; the data analyzed in this article are limited to those households that accepted the offer of affordable housing and continued to live in the complex through the time of interview.8 We recruited the head of household, defined for the purpose of this study as the individual who completed the initial housing application. In some cases, the head of household was unavailable, was not English proficient, or preferred not to be interviewed. For these households (N = 7), we recruited another adult member of the household if that person was part of the original household that moved to the study site (that is, was listed on the initial housing application). The response rate was 64 percent.9 Face-to-face interviews were conducted in the home, at the project’s offices, or at another location based on the preference of the respondent. Interviews lasted approximately 50 minutes and included a series of name-generator and name-interpreter loops to create the Additional affordable housing units in these two buildings were allocated to eligible households that did not apply through the housing lottery.
In the present analysis, we exclude 18 households that no longer lived at the affordable housing complex.
Of the 241 households, 18 were defined as out of scope because of language (that is, the householder was not English proficient and no other adult household members were eligible). An additional 4 households were deemed out of scope for the present analysis, including 1 that was unable to provide informed consent and 3 in which the household member who was interviewed was not on the original housing application. Another 14 households had unknown eligibility status.
In these cases, the identity of the household could not be confirmed for reasons such as a language barrier, no contact established after several attempts, or the householder no longer lived in the sampled unit but could not be confirmed as living somewhere else in the complex or having moved elsewhere. The final response rate is calculated using the American Association for Public Opinion Research Standard Definitions, Response Rate 5 (AAPOR, 2011), which excludes ineligible and unknown eligible households from the denominator. Our final response rate is calculated as 120 completed interviews / (120 completed interviews + 1 incomplete interview + 66 refusals) = 64 percent.
egocentric network data analyzed in this article. All interviews were conducted in English.10 All protocols and materials for this study were approved by the Institutional Review Board at Teachers College, Columbia University (Protocol #12-175).
Interview data were linked to additional secondary data collected before move-in, which were used to describe the population served. Baseline data were obtained via a self-administered questionnaire completed before the final determination of eligibility for housing (87 percent of the respondents analyzed in this article also participated in the baseline survey), via self-report information obtained from the housing application, and via other data collected by the housing developer as part of the screening process (administrative data were available for all 120 participating households). Exhibit 1 presents basic descriptives of the study population. Affordable units include studio, one-bedroom, and two-bedroom units, with mover households ranging from one to four people. At the time of the interview, 37 percent of the households had one or more coresident children and 29 percent were single-person households. Most respondents were female, with a median age of 40 at the time of interview. Overall, this population is educated, with 49 percent completing a 4-year college degree or beyond. At the time they were interviewed, 76 percent of respondents were working for pay; the median household income was $45,000.11 As a result, 18 households were defined as out of scope because they were not English proficient.
Employment status was not collected for other adult members of the household; therefore, it is likely that a greater proportion of households has at least one wage earner than reported here.
Baseline data come from administrative data confirmed by the housing developer as part of the housing eligibility screening b process.
Baseline data come from self-report information contained in the initial application for housing.
c Using fiscal year 2009 U.S. Department of Housing and Urban Development Income Limits for the New York Metro Fair d Market Rent area.
Baseline data come from self-report on baseline self-administered questionnaire (SAQ). Percent shown as proportion of total e completing SAQ (N = 103).