Abstract
While supportive social ties help to buffer against the consequences of poverty, few researchers have examined how people form such ties. New ties are often formed in routine organizations such as businesses, churches, and childcare centers, which, beyond being places to work, shop, or receive services, are institutionally governed spaces of social interaction. Based on the notion of organizational brokerage, we introduce a perspective that specifies when routine organizations contribute to tie formation and use it to reexamine data from existing qualitative studies of such organizations among the poor. We argue that successful brokerage will depend on the degree to which an organization’s institutional norms render interaction among participants frequent, long-lasting, focused on others, and centered on joint tasks; and that the ensuing networks may differ from other supportive ties in the sense of belonging they may cultivate, the form of generalized exchange they may engender, and the organizational connections they may create.
A large literature has shown that social networks are essential for the ability of low-income populations to buffer against the consequences of poverty (Stack 1974; Nelson 2000; Domínguez and Watkins 2003; Small 2009; Raudenbush 2016; Desmond 2012). In fact, the literature is so large that one could be forgiven for believing that being poor somehow provides automatic access to a network of supportive social ties (Nelson 2000; see Smith 2007). However, many low-income people do not have such a network (Campbell, Marsden, and Hurlbert 1986). In fact, survey-based studies have found that poorer people living in high-poverty neighborhoods have smaller nonkin networks and are more likely to be isolated than those living in low-poverty areas (see Small 2007; van Eijk 2010; Burdick-Will 2018; Soller et al. 2018; see also Offer 2012). How, then, do low-income people form networks of support?
We examine the role of the organizations in which people interact with others on a routine basis as part of daily life: organizations such as workplaces, churches, childcare centers, schools, soup kitchens, gyms, bars, neighborhood restaurants, community centers, and other establishments (Oldenburg 1989; Small 2009; also Hsung, Lin, and Breiger 2009; Mollenhorst, Völker, and Flap 2008; Mazelis 2017; Klinenberg 2018). Though routine organizations are arguably the primary means through which low-income—and other—populations form ties outside the family, there are few systematic theories about how this process works; about why people form new ties in some contexts but not others; or about how, if at all, the ensuing relations differ from family or other ties in their supportiveness (but see Small 2009). We develop a theory of how people, regardless of their income, form social ties in such organizations; probe its applicability based on a reading of published U.S. field studies in the literature on the poor; propose at least four factors that distinguish routine organizations in which people are likely to form ties from those in which they are not; and identify several ways organizationally brokered ties may differ from other supportive ties.
Our Study
Our study is motivated by substantive, theoretical, and policy concerns. The substantive motivation is the arguably rising importance of nonfamily support networks. Support networks are valuable to individuals at all points in the income distribution. Yet the last two decades have heightened the need for low-income families in the United States to secure social, economic, and practical resources from their networks. Over this period, cash assistance has decreased for low-income mothers of young children, part of an ongoing restructuring of the U.S. welfare state (Moffitt 2015, 742–43). The Personal Responsibility and Work Opportunity Reconciliation Act (PRWORA) of 1996 dramatically reduced welfare rolls and made the poorest nonelderly families increasingly have to fend for themselves. Subsequent studies of how people avoid homelessness, material hardship, and other difficulties repeatedly found that social networks, especially family networks, were important (Domínguez and Watkins 2003; Watkins-Hayes 2013; Harvey 2018; see also Newman and Massengill 2006). Yet such conditions have strained many family networks; for example, many low-income parents have been forced to move in with adult relatives (Harvey 2018). As cash assistance has subsided and employment has become more unstable, people increasingly turn to their networks for support. Yet the strain on family has become substantial, heightening the importance of forming nonfamily ties.
The theoretical motivation of our study is the need for clearer theories of how organizational processes play a role in both urban poverty and social inequality (Marwell 2007; Small 2009; Wacquant 2009; Sampson 2012; Avent-Holt and Tomaskovic-Devey 2019). While research on urban poverty has traditionally focused on either the individual or the neighborhood, researchers have recently rediscovered the role of organizations in many aspects of the urban condition, including how concentrated poverty affects life chances; how people decide where to live and work; and how information, goods, and other resources are distributed across networks (Marwell 2007; Allard 2009, 2017; Small 2009; Sampson 2012; Allard and Small 2013; see Galaskiewicz and Marsden 1978; Warren 1978). A similar recognition is evident among students of social inequality, who have argued that organizations are central to differences in resource access, to the development of status distinctions, and to the functioning of relational inequities (Tilly 1998; Small 2009; Avent-Holt and Tomaskovic-Devey 2019). Organizations are equally central to the formation of social ties among urban poor populations, and the processes through which they operate remain undertheorized and poorly understood.
The policy motivation is the need for clear guidance for practitioners and legislators who seek to intervene at the level of the network. Though U.S. policy-makers have frequently invoked the importance of social networks and social capital to the survival of low-income families, they remain uncertain on how to incentivize or otherwise help people to build those connections (see, e.g., Sommer et al. 2017). After all, just asking people to make more friends is not an effective strategy. Furthermore, such ideas can come close to “blaming the victims” (Ryan 1976, xiii), given that conditions that often result from structural factors may seem to result from inadequate networking. Organizations such as childcare centers and community centers can be effective places to intervene, provided one knows how to do so. Yet the research has produced precious few models for how practitioners would intervene in local organizations to make them more effective places to build such ties.
In what follows, we address these needs by examining when routine organizations help low-income people to form new social ties, identifying the mechanisms through which it happens, and documenting the operation of these mechanisms based on an extensive literature review of U.S.-based case studies of routine organizations frequented by low-income families. We build on network theories about the importance of brokers as conduits between previously unconnected actors and suggest that the process of meeting others in routine organizations results from organizational brokerage (cf. Small 2006). Organizations, not just individuals, can connect people to others.
We argue that routine organizations can broker social connections via multiple types of mechanisms, which include those driven by actors and those driven by institutional practices. We focus primarily on one kind of institution-driven mechanism—the impact on social interaction among participants—and argue that successful brokerage will depend on the extent to which institutional norms render interaction frequent, long-lasting, focused on others, or centered on joint tasks. We suggest that the ensuing networks may differ in important ways from other supportive ties. We conclude by outlining a set of issues that remain unanswered.
A Perspective on Tie Formation
Previous work
The literature on how actors form ties to others is vast, diverse, and multifaceted 1 (e.g., see Lazarsfeld and Merton 1954; Newcomb 1961; Verbrugge 1977; Feld 1981; McPherson, Smith-Lovin, and Cook 2001; Rivera, Soderstrom, and Uzzi 2010). We cannot review it all. For our purposes, the most relevant body of work has taken the individual as the unit of analysis and conceived of tie formation as the result of two separate processes, “meeting” and “mating,” or coming into contact with others and then deciding to associate (Verbrugge 1977). The literature has isolated the meeting process to examine the factors that affect the opportunities for social interaction. As Blau and Schwartz (1997) put it, “Rates of social association depend on opportunities for social contact. . . . [T]he extent of contact opportunities governs the probability of associations of people, not merely of casual acquaintances but even of intimate relations, like those of lovers” (p. 29; see also Blau 1977, 90; Marsden 1990, 397; Gans 1961). For example, researchers have examined at length how organizations and other “foci” bring unconnected people into contact (Feld 1982; Small 2009). The literature has isolated a second process to examine the factors affecting the decision to form a tie. As Marsden (1990) put it, from this perspective, “variation in network composition [is] the result of differing levels of individual preference for associates of particular kinds” (p. 397).
Dividing the process into the opportunities to come into contact and the decisions to associate given that contact is powerful and useful. But it has an important limitation. When applied to our current context, what factors affect the probability that a person will patronize, say, a barbershop (providing a chance to meet), and then separately the process through which they decide to associate (choosing to mate)? The problem with this perspective is the assumption that once people come into contact with others, it is merely up to them. For example, many people go to barbershops and connect with no one; that outcome, from this perspective, would have to be a function of their low self-efficacy, extraversion, motivation, friendliness, or general inclination to make friends. Such an account is implausible, since equally motivated or efficacious people can patronize different barbershops and differ in the rate at which they interact with others, depending on the barbershops’ environments. In fact, Small (2009) has reported such differences among equally motivated parents who patronize different childcare centers, with some expanding their networks meaningfully and others not doing so. As we document below, the conditions of the organization can dramatically affect the degree of network formation among people who have already had the opportunity to come into contact. Presuming that the second, postmeeting part of the formation process depends only on individual decisions is an undersocialized understanding of human actors (Wrong 1961). In sum, the distinction between contact and choice, or between the opportunity to interact and the decision to befriend, elides the important mediating roles that organizations can play even after the “meeting” process has taken place.
A different perspective
We propose a theory that begins where the “meeting” process ends. Rather than presuming that, once contact has been established, only the decisions of the individual matter, we assume that the postcontact process is shaped by the organizations in which contact took place. We take for granted that agency matters, that some people are more efficacious than others, and that highly efficacious people will, except in extremely inhospitable contexts, essentially introduce themselves to others (see Emirbayer and Goodwin 1994; Emirbayer and Mische 1998). Even among such actors, organizations can make the process easier or more difficult, so the mediating role of organization must be understood.
A mediator is a broker, and our model builds on brokerage theory. A broker is traditionally defined as the link between two unconnected actors (Simmel 1950, 1955; Gould and Fernandez 1989; Burt 1995, 2005). As Stovel and Shaw (2012, 141) write in a recent review, “Brokerage [is] the process of connecting actors in systems of social, economic, or political relations in order to facilitate access to valued resources.” While literature has examined brokers as tertius gaudens (Simmel 1950; Marsden 1982; Burt 1995), people who gain advantage as a result of their position, we instead explore brokers as tertius iungens, those who, because of their location, bring others together (see Obstfeld 2005). Our model of how brokers join others relies on three core assumptions: that brokerage is a process, that the process is organizational, and that the organizational effect is multifaceted.
Brokerage is a process
We assume that brokerage is a process. The majority of research on brokerage in network analysis has focused on the structure of relations that give brokers their advantage or opportunity to connect others. In contrast, Obstfeldt, Borgatti, and David (2014) argued persuasively that brokerage is not merely a structure but also a process. Brokers must do things to connect people, and the things they do can be important. In fact, focusing only on structural conditions may miss that the things a broker does to connect two people may be more important than whether those people had never met before or only had a passing acquaintance (and were, thus, not technically unconnected). Conversely, the things a potential broker fails to do may result in a missed opportunity, regardless of how well positioned in a structure the broker is. As the authors argue, brokerage theory must take into account “the social behavior of brokering” (Obstfeldt, Borgatti, and David 2014, 139). Consistent with this prior research, we focus not on formal structure but on the process through which brokerage happens.
The process is organizational
We assume that brokerage may be affected not merely by individuals but also by organizations. We define a routine organization as a space of interaction where a set of actors, guided by institutional norms and understandings, orient their activities and practices loosely toward a global purpose (see Powell and DiMaggio 1991; Scott 1995; Small 2009). Routine organizations include workplaces, restaurants, barbershops, childcare centers, grocery stores, bodegas, churches, botanicals, gyms, community centers, neighborhood clinics, coffee shops, bars, and so on. Though routine organizations vary in scope, mission, profit status, funding source, orientation, and many other conditions, they all constitute a space where actors interact around institutionally shaped norms and understandings. 2
Our view that organizations can operate as brokers differs in important ways from the existing research. To date, researchers have understood the role of organizations as brokers in one of two ways. One, they have conceived the organization as an actor whose behavior is analogous to that of an individual. For example, an organization, not just an individual, might be said to gain from its brokerage position, thus having greater power to set prices (see, e.g., Marsden 1982; Stovel, Golub, and Milgrom 2011; Stovel and Shaw 2012). From this perspective, the organization and the individual are two instances of the same general class, and analogous processes occur in each subclass. Two, researchers have conceived the organization as a community, equivalent to any entity comprising members, including groups, parties, associations, affiliations, or even identities. For example, any community might be said to affect how a broker performs a role depending on whether the broker is a member (and thus, say, a “representative”) or not (and thus, say, a “liaison”) (see Gould and Fernandez 1989). From this perspective, organizations are analogous to any membership entity that might affect whether a broker is an insider or an outsider—a broker who is Christian is a member of the community of Christians vis-à-vis any non-Christian, a Democrat broker is a member of the community of Democrats vis-à-vis Republicans, and so on.
In contrast, we assume that organizations are analogous to neither individuals nor communities. Though routine organizations range widely in type and orientation, all constitute a space of interaction, which cannot be said of individuals or all other kinds of communities. As a result, some forms of brokerage are distinct to organizations.
The organizational effect is multifaceted
We assume that organizational brokerage may occur through not one but multiple types of mechanisms. An organization is a space of interaction, a set of actors, and a constellation of practices, all shaped by institutional norms and understandings. Indeed, the difficulty in developing a clean conceptual model of how they shape networks is that people can create ties as a result of multiple processes. There are multiple ways an individual who enters a space, who interacts with actors, or who responds to norms or understandings may come to meet others whose resources prove useful to their survival.
Actor-Driven Brokerage
One way to systematize these various processes of interaction is to focus on the entity doing the brokerage. Sometimes, it is just a person, and the process is a straightforward application of brokerage theory. Any time an individual interacts with an acquaintance in an organization, that acquaintance may introduce the individual to others. In that case, the acquaintance would be acting as tertius iungens, precisely as described by previous researchers, merely in the confines of an organization. Just as in traditional brokerage (which we will call “type A”), the process is being driven by an actor who is a member of the organization. For our purposes, an organizational member may be a manager, employee, client, patron, or volunteer of the organization—since the act of brokerage may be perpetrated by anyone in the organization. 3
This actor-driven brokerage process can take at least two different forms, depending on whether the broker is connecting the individual to another member or to a nonmember of the organization (see Figure 1, panels A, B, and C). For example, a church pastor may connect a parishioner to a pastor at a sister church (type B) or to another parishioner in the same church (type C).
A Model of Organizational Brokerage
Both of these types of brokerage have been documented in research on routine organizations in urban contexts. Desmond’s (2012) fieldwork in Milwaukee uncovered that low-income people made connections in “welfare offices, food pantries, job centers, Alcoholics Anonymous clubs, methadone clinics, even the waiting areas of eviction court” (p. 1313). One of Desmond’s cases illustrated vividly the type-B organizational brokerage seen in Figure 1: Scott and Mike met while attending a Cocaine Anonymous meeting. Mike introduced Scott to Pito, a contact of Mike’s outside of the support group, who then introduced Scott to David. After Scott was evicted, he ended up rooming with David (p. 1312).
Researchers have also documented the type of simple brokerage in which individuals introduce unconnected actors in the same organization (type C). This type of brokerage is particularly evident in schools, as Lukasiewicz and her coauthors (2019) found after studying social capital among low-income parents in New York City. One parent served as the president of the Parent Teacher Association (PTA) and often brokered relations between parents and between parents and staff. She explained that her role was especially important for parents who did not speak English: “They could come to me and I will speak for [them]. So it has helped, it has changed the school. . . . And we have done a lot of events for the school, for the kids, that wasn’t going on for a couple years” (p. 287).
In both cases, the process is just conventional brokerage; nevertheless, it is rightly thought of as organizational, because in the absence of the organization, the brokerage would not have happened—Scott would not have met Pito, and the non-English-speaking parents would not have communicated with others. The organization placed Mike and the PTA president in a position to connect others. 4
Institution-Driven Brokerage
Organizational brokerage may be driven not only by actors but also by the organization itself. An organization can structure the formation of ties among its members in a large number of ways, including its rules of membership or participation, its physical layout, and its institutional norms or understandings of social interaction (see Small and Adler 2019). We focus on the latter, on routine organizations’ institutional practices. Practices may be “institutional” in the normative sense—via formal rules or informal norms that encourage or discourage forms of behavior; or in the cognitive sense—via understandings of the self or others that are produced by the organization (Small 2009; see Scott 2003). An organization may encourage practices through which members meet people in different contexts (see Figure 1, panel D), as when clinics send health workers on mobile outreach operations to serve families in low-income neighborhoods. Or an organization may have rules of social interaction that encourage people to meet other members (see Figure 1, panel E), as when schools expect parents to collaborate on fundraising activities. In such cases, no individual is actually doing the brokering.
Researchers have also documented cases of low-income individuals meeting others through organizations despite that no actor is introducing one actor to another. In such circumstances, the institutional practices of the organization have been key. The first of these, type D in Figure 1, has been reported in different forms. In his Boston study of twenty-nine religious organizations in low-income neighborhoods, McRoberts (2003) found that the churches “are not just places where people meet to worship”; instead, they also acted as “interactive social spaces and as architects of vertical and horizontal networks,” helping congregants form connections within and between churches (p. 127).
An examination of type-E brokerage
The bulk of our discussion, however, is on type-E processes. We examine how institutional norms shape four aspects of social interaction among members of an organization: how repeated it is (frequency), how long-lasting it is (duration), how focused it is on others (outward orientation), and how centered it is on the accomplishment of joint tasks (collaboration). These factors can dramatically affect whether a routine organization will successfully broker connections.
Frequency
Repeated interaction is the foundation of friendship formation (Homans 1961; Lazarsfeld and Merton 1954). Such interactions can lead to the discovery of similarities and the subsequent operation of other mechanisms such as homophily (McPherson, Smith-Lovin, and Cook 2001). Local routine organizations shape the support networks of low-income families to the extent that they allow, encourage, or require repeated visits that serve as opportunities for relationships to form between individuals (Small 2009).
Research on churches has often documented this relationship, given the repeated nature of church participation. Ellison and George (1994) examined the relationship between church participation and network size based on North Carolina survey data (n = 2,956). They found that respondents who reported attending church “several times a week” had, on average, 2.25 more nonkin ties than those who “never” did; and the more they attended, the more nonkin ties they had (p. 56; see also Taylor and Chatters 1988). In her fieldwork in a mixed-class African American neighborhood in Chicago, Pattillo (2013) found that many young people developed strong bonds to others through weekly choir rehearsals at a local Catholic church (see also Marwell 2007).
Likewise, the daily requirements surrounding childcare make for multiple opportunities in which parents and childcare providers can interact (see Small 2009; Reid, Martin, and Brooks-Gunn 2017). Small found that “practices such as constrained pickup and drop-off times and monthly meetings with teachers facilitated frequent interactions,” which contributed to tie formation (2009, 103). Researchers have also documented the formation of other kinds of ties that provided social support. For example, Rolfe and Armstrong (2010) examined communication with early childhood professionals among mothers at their child’s daycare center and found that thirty-eight of the forty surveyed mothers reported speaking in person with their early childcare providers at least three times a week. Most of them perceived the early childhood professionals to be a “source of social support”; furthermore, the frequency of interaction “determined how much support mothers considered they received” (pp. 62, 60; see also Hughes 1985).
Research on community centers has documented this idea as well, given that such centers often support programs that require participants to meet regularly. Colistra and colleagues conducted a case study with fourteen program participants and found strong evidence of this process shaping tie formation and trust (Colistra, Schmalz, and Glover 2017; Colistra, Bixler, and Schmalz 2019). As one parent whose child attended the after-school program explained, the daughter “became close” to the other participants because she was “seeing them every day” (Colistra, Bixler, and Schmalz 2019, 10; see also van Eijk 2010).
Another context in which researchers have found this relationship is hair salons and beauty salons, given the routine nature of interactions in the space (see Furman 1997; Harris-Lacewell 2004; Shabbazz 2016). Exploring the social relationships formed in a black barbershop on Chicago’s South Side, Harris-Lacewell (2004) found that major social connections were forged through frequent and extensive conversations in the shop—“authentic everyday talk” (p. 201)—which, notably, extended beyond class lines.
Soller at al. (2018) interviewed 178 recently resettled refugees in a large southwestern city to understand their community attachment and social networks. The authors asked respondents about the number of times they visited specific “grocery stores, restaurants, and places of worship in the last thirty days” in the community (p. 337). Refugees who engaged more at local stores and organizations had a more robust local network. The authors concluded that “members of ethnic communities who share interactional settings with other members when engaging in their routine activities are more strongly attached to other members and have enhanced access to resources that are embedded within their communities” (p. 340).
Duration
A second factor that plays a role in brokerage is the presence of institutional norms, which allow people to spend extended periods in the company of others. The passage of time contributes to sociability (Simmel 1950), the possibility of conversation with no instrumental purpose other than the conversation itself, a process through which strangers become acquaintances, and acquaintances friends. Research on “third places,” the bars, pubs, beer gardens, and other establishments where people can socialize at length without being mindful of the clock, has often made this point (Oldenburg 1989,17).
Research on neighborhood restaurants has confirmed this idea. Duneier’s (1992) study of a neighborhood cafeteria on Chicago’s South Side illustrated the extensive community that can form when clients can patronize an eating establishment for long periods. The cheap prices and lack of expectation for patrons to depart upon payment encouraged many to stay for hours and interact with one another throughout the day. Over months of observation, Duneier found that the men of Slim’s Table passed time at Valois by “participating in the same rhythm of various routinized episodes that yield both companionship and solitude” over the many hours and days regulars spent there (p. 34). In fact, the “ties binding members of the larger collectivity [at Valois] have developed over decades” (p. 5).
Research on neighborhood bars suggests the same (e.g., May 2001; Anderson 1978). Anderson’s (1978) study of a neighborhood bar and liquor store on the South Side of Chicago examined the social interactions that emerged among regulars, who spent hours at a time at the bar. The relations formed were so significant that the author termed them, following Cooley (1909), an extended primary group, and described them as possessing a “we” feeling (Anderson 1978, 33) in which the group provided “supportive social ties for its own” (Anderson 1978, 180).
In the community center Colistra and colleagues studied, seniors met for four hours during each programming session. The results were as expected. As Barbara explained, “I think it is really the program what brings us all together, when we join in and do things together as a group, and that connects us you know” (Colistra, Bixler, and Schmalz 2019, 10).
We note that in this and other forms, routine organizations may differ widely not only across types but also within them. For example, contrary to the one Duneier (1992) studied, most restaurants in the United States do not permit patrons to linger for hours after eating, lest it cut into the businesses’ often slim profit margins. The difference across entities of the same kind is part of why we focus not on organizations as “types” but on the mechanisms through which they shape interaction.
Outward focus
In a series of important papers, Feld (1981, 1982, 1984) proposed that social ties were formed when people interacted around a common focus of activity. This “focused interaction” could happen in multiple ways and across many kinds of foci. In an organization, the institutional norms or understandings may encourage interaction to be less or more focused in nature. The interaction may be focused in at least two ways, as we describe in this section and the following one.
The first is the extent to which the activity of any given member is oriented toward others, rather than themselves. Orienting one’s activities toward others contributes to tie formation through several processes. It naturally encourages social interaction; it helps people to identify common values (Lazarsfeld and Merton 1954); and it facilitates their ability to discover similarity or homophily in background (McPherson, Smith-Lovin, and Cook 2001). These conditions, in turn, help actors to find points of cognitive empathy, the ability to understand another’s predicament from their perspective, which builds trust (Small 2017). An organization’s institutional norms may encourage this kind of outward focus, particularly as a result of purposeful group engagement.
Researchers have documented such practices among organizations patronized by low-income populations, including churches, community centers, and service agencies. Based on ethnographic fieldwork in Brooklyn, Marwell (2007) found that a major church managed to develop especially strong relations among congregants through its fraternidades—communal living groups centered on spiritual practice and religious connection through prayer, reflection, community service, and chores. Members of the fraternidades were expected to participate in revisión de vida sessions, intimate group gatherings where participants shared personal stories with one another and reflected on their “daily lives” (p. 193). Watkins-Hayes (2013) interviewed eighty-two HIV-positive women who attended weekly group meetings at an AIDS service organization. As part of the meetings, participants were expected to share their experiences and listen constructively as others described theirs, encouraging each to focus on the others. The shared experiences produced a natural environment for network formation, emotional support, and advice (p. 91). As a participant explained, “I’m just outreaching for anyone because I’ve been there and I know sources. . . . I don’t mind plugging nobody in because we’re all in this together” (p. 92; see also Watkins-Hayes 2019).
Community centers can have a similar effect, as many of their programs require participating in collective sessions. Bess and Doykos (2014) studied a “cradle-to-college” program for parents and caregivers in an East Nashville community center (p. 272). Participants were required to attend multiple kinds of activities, including those during which family shared personal or community news with the others. The authors found that the majority of program graduates had formed new relationships to other participants (p. 275). Similar kinds of programs are often run either at or by schools. Shoji et al. (2014) interviewed parents and ran focus groups with participants in one such program aimed at “empowering” parents and increasing their involvement at the children’s school (p. 602). The parents met weekly for eight weeks and were expected to participate in group discussions, where they shared aspects of their predicament. Supportive social ties were formed, as expected. Activities of this kind are particularly common among organizations with a service orientation.
Joint tasks
A second way the activities in an organization may be focused is the extent to which people collaborate to perform concrete tasks. Research in social psychology has uncovered that positive affect and trust may increase when strangers collaborate on a joint project (Lawler 2001). To work successfully with others, people must represent their best selves, adopt a collaborative attitude, and communicate. To the extent institutional norms require collaborations among members to meet organizational needs or objectives, they will encourage the formation of social ties (Small 2009).
Research among low-income populations has also uncovered this relation. Several studies have focused on childcare centers, which often expect parents to work on activities to meet collective ends. In a study based on survey and ethnographic research from childcare centers, Small (2009) found many centers expected parents to collaborate with one another to accomplish major tasks. Parents may be expected to help organize fieldtrips; to plan special events, such as Mother’s Day celebrations; to arrange spring cleanings of playgrounds; or to put together fundraising activities to meet annual targets. These activities consistently contributed to the formation of social ties. Reid, Martin, and Brooks-Gunn (2017) found something similar based on interviews with parents associated with childcare centers in New York City. They found that when parents engaged in center activities such as open houses and conferences, their interactions with one another and with center staff increased, promoting the exchange of information and emotional support (p. 148).
What Resources Do These Ties Provide?
One might imagine that the ties low-income individuals form in such organizational contexts are somehow less valuable than those formed otherwise, that the formal settings in which they were forged perhaps lack the intimacy needed for truly dependable relations. The evidence is not consistent about this. In fact, organizationally brokered ties are not only useful in the ways other ties are; they are valuable in ways distinctive to organizationally embedded ties. We briefly identify three conventional resources formed through organizational brokerage and three different kinds of resources that derive from the organizational nature of the ties.
Emotional support
Organizationally brokered ties have been documented to provide emotional support. Klinenberg’s (2018) survey- and interview-based study of libraries and other local organizations found these establishments useful in “buffering isolation and loneliness” (p. 34). Denise, a young mother who frequented her local library, explained that she found a lot of support from other mothers and babysitters. “You just kind of start chatting . . . and it’s amazing, but you wind up having these really personal, really intense conversations” (p. 35). Using interview and observational data at a youth-centered community technology center, Clark (2005) found cases of bridging social capital, in that activities at the center, like decorating a Christmas tree, brought together participants from different neighborhoods (p. 437). The center offered youth participants a space in which to talk openly and freely about issues at school or home with one another, particularly because of the openness of the space. Participants talked about issues such as sex education, how to deal with bullies, and family life. Using nationally representative data, Small (2009) found that mothers who enrolled their children in childcare centers and formed ties were substantially less likely to experience depression than those who enrolled but did not form ties or than those who did not enroll, after accounting for prior depression.
Information
Organizationally brokered ties are also good sources of information. Colistra, Schmalz, and Glover (2017) examined social relationships and well-being at a county-owned community center through interviews and focus groups with fourteen adults who frequently visited the center. The authors found that social relationships supported well-being through the “exchange of resources, services, and information” (p. 42). For example, Brenda sought legal assistance from other community center members who worked as lawyers: “We have lawyers in programs and things here so I get to reach out to them and pose my little question to them and they will guide me and let me know how to work things out with whatever the problem might be, especially my house. But, you know, I’ve met a lot of people who have helped me since I have been here, and especially in the legal field” (p. 42; see also Delgado 1996, 1997, 1998; Delgado and Santiago [1998] on shops; Johnson 2010; Lopez, Caspe, and Simpson 2017; Khoir et al. 2017; Klinenberg [2018] on libraries).
Services and material goods
These ties have also been sources of more concrete forms of support, including services and material goods. In the community organization studied by Shoji and colleagues (2014), parents expressed the value of the new social ties in terms related explicitly to the services they offered. One parent explained, “When my baby was born, I couldn’t go on Fridays because of my baby boy, and I could tell one of them [the other parents in the program], ‘Can you pick up my daughter? Can you bring her home?’ You know what I mean? And if I hadn’t met them, it would have been more trouble” (p. 607; see also Lukasiewicz et al. 2019). In her study of a social services assistance and advocacy group in Philadelphia, Mazelis (2017) found repeated evidence of resource exchange. One member, James, explained, “People always been helping me since like I got involved with [the organization]. I done met different people in different relationships that allowed me to be able to function, me and my [immediate] family. From transportation stuff to money stuff to food stuff, to you name it, people that helped me over the last seven years” (p. 155).
Generalized exchange
Some resources derive from the organizational context in which the ties are created. One such resource, which results because ties can take collective form, is the security that derives from generalized exchange. We cannot do justice to the vast literature on the topic here, but we note that researchers have defined generalized exchange as a system in which the receipt of goods or services may be reciprocated by someone other than the receiver (Malinowski 1922/2014; Blau 1964/1986; Molm, Takahaski, and Peterson 2000). When ties are formed in a group context, and particularly when they stem from collective activities, their relationships may naturally evolve into this kind of system.
In the community center studied by Colistra, Schmalz, and Glover (2017), participants often described such systems. Linda explained, “like when I needed food, transportation . . . and you know, [other members] don’t just direct me, they will help me themselves, they will give me something, or ‘here [participant’s name] take this home’” (Colistra, Schmalz, and Glover 2017). Similar systems can emerge even in places that might not seem “community-focused,” such as bars. Anderson (1978) reported such relations in the bar that he studied, where extended socialization over drinks helped the regulars to feel “themselves among equals, especially in relation to wider society” (p. 29). When Herman, one of the more prominent regulars, was showing up infrequently and run down because his wife had fallen ill, other regulars asked their wives to pack food for him or otherwise brought extra “grub” from their restaurant jobs for him (p. 181).
Mazelis (2017) reported a similar process based on interviews and fieldwork in a Philadelphia organization that helps low-income families obtain housing and social services. To maintain membership and receive aid, the group required members to volunteer their time at the local office, in assisting others, or at rallies. Such expectations ultimately encouraged generalized reciprocity. One member, Cate, learned to navigate the welfare system and reported increased self-esteem: “[being a member] feels like you’re secured. People are securing you. . . . Like before it’s like I felt doors was locked. When I feel that people’s helping me, doors are opening” (p. 128). To fulfill the expectations for giving back, Cate lets “new members stay with her, travels with the group, goes on food distributions, and brings in other members” (p. 128).
A sense of belonging
The collective nature of much of tie formation can also create, at times, a sense of belonging to a group larger than oneself. In May’s work (2001), people found their engagement with others at a neighborhood bar to contribute positively to their self-worth. May concluded that the “exclusionary” nature of the tavern created an “environment where a regular clientele affirms its own racial or ethnic identity” (p. 172). According to May, for some clients the experience of conversation was akin to visiting a therapist, as the bar was a safe space where patrons could “claim positive identities for themselves” (p. 163).
Researchers have made similar claims about barbershops (Shabazz 2016). Men who frequented the barbershop Shabazz observed over a summer found social opportunities to learn “about black history, developing skills in argumentation and debate, and male bonding,” building a sense of group identity (2016, 310). Rock, a frequent customer, explained, “I used to take my son with me all the time so he could soak the knowledge. You can’t get that love anywhere else” (2016, 306). The engagements within the barbershop give clients, particularly those who are younger, a space in which to develop ideas and arguments and to learn more about their history and their own identities.
Access to other organizations
Finally, some of the ties formed are not merely to other people (as in B and D) but to other organizations. In Colistra, Bixler, and Schmalz’s (2019) aforementioned study, the community center “collaborated with other community centers to deliver programs, thus connecting people from different neighborhoods” (p. 11). Center-user Barbara explained, “We get invited to different centers to join them for different functions they have and two or three times out of the year, and then we invite other centers here to interact to join in with us” (p. 11). These connections to centers, churches, or other entities provided their own resources and opportunities to combat isolation. Another user explained, “We have a choir of seniors and they get invited to [local church], and they have been to [the community center] to sing,” and there are “some people who stay very closely contacted and connected with seniors in other programs” (p. 11). McRoberts’s (2003) study of Boston-area churches found multiple organizational ties between one church and other community organizations. One church “heavily involved” their congregants in outreach work (p. 118), which included “a night patrol aimed at establishing contact with youth in the street, a number of educational programs for young people and their parents, and the Four Corners Planning Committee” (p. 126).
Conclusion
We have developed a perspective on tie formation in which organizations are conceived as brokers and their institutional practices as the mechanisms through which brokerage occurs. We have applied this perspective to understand and document tie formation particularly among U.S. low-income families. We have argued that successful brokerage will depend on the degree to which institutional norms render social interaction frequent, long-lasting, outwardly focused, and centered on joint tasks; and that the ensuing networks may differ in important ways from other supportive ties.
We take our article to be a starting point; many questions remain unanswered. Three of them are particularly important. One question relates to the role of agency, preferences, and decision-making in how people respond to institutional conditions. While we have focused on how institutional conditions shape behavior, people are agents in their network formation process, and whether they ultimately create a tie depends on their decision to do so (Emirbayer and Goodwin 1994; Emirbayer and Mische 1998). For example, people will likely seek those who resemble them demographically (Lazarsfeld and Merton 1954; McPherson, Smith-Lovin, and Cook 2001; Wimmer and Lewis 2010). In addition, the friendship formation process is dynamic in ways likely to matter. For example, people will differ in how much they invest in relations they have begun to form, thereby affecting whether fragile new ties become strong or dissipate. Similarly, they may or may not decide to sustain a relation once the initial brokerage provided by the organization has taken place. The interactions between agency and context remain a fruitful area for new work.
A second set of questions involves the factors motivating organizations to serve as brokers. As we discussed, the motivations, interests, and benefits to brokers have concerned researchers for many years (Burt 1995; Stovel and Shaw 2012). The organizational dimension does not eliminate those concerns; it merely complicates them. The process through which the routine organizations shaped tie formation in the cases we discuss did not necessarily derive from an intention to form connections among patrons or members; in fact, in most cases, they derived from other organizational imperatives (Small 2009). The variations in intentionality and objectives, and the unintended consequences that ensue, deserve substantial attention. This set of issues is vast and represents a particularly rich space for new research.
A final set of questions involves variation in the types of organizational membership that individual actors may have. Our theoretical perspective identifies three kinds of actors: the focal one, the broker, and the person connected by the broker to the focal actor. In strict theoretical terms, any of them may be organizational members, and such membership may take any form: client, patron, manager, employee, volunteer, and so on, as long as the form involves interaction in a space with other members. Because of length constraints we have been forced to ignore the consequences of this heterogeneity. And because of the article’s focus on low-income populations, and the cases the literature has chosen to explore, we have primarily documented scenarios where the focal actor and the connected person are both usually a client. This focus has left many questions unanswered. Indeed, panels B, C, D, and E in Figure 1 are all affected by these conditions. Even in those circumstances where the focal actor is a client, whether the connected person is a client versus an employee carries vastly different implications for how the brokerage process will ensue, for how likely the relation is to be maintained, and for what kind of social support resources are available through it. In this respect, our application has only scratched the surface of the questions needing examination.
We believe that these and other questions can help to inform an agenda in which research on poor populations focuses not only on the individual or the neighborhood but also on the organization. This agenda also provides a clearer path for practitioners and policy-makers. The effort to stimulate network formation among low-income populations has reason to work through local organizations. The particular mechanisms that we have identified provide a potential starting point.
Footnotes
Note:
We thank Miranda Lubbers, Hugo Valenzuela, and the anonymous reviewers for comments and criticisms that have improved this article. We acknowledge the generous support of Harvard University’s Faculty of Arts and Sciences; its Business School; and its Project on Race, Class, and Cumulative Adversity, at the Hutchins Center.
Notes
Mario L. Small is the Grafstein Family Professor of Sociology at Harvard University and is an expert on poverty, inequality, ego networks, and field methods; his most recent book is Someone to Talk To: How Networks Matter in Practice (Oxford University Press 2017).
Leah E. Gose is a doctoral student in sociology at Harvard University studying community organizations and social service provision.