Beyond Their Absence: Male Intergenerational Social Ties and Community Informal Social Control

Background

Informal social control, or the behaviors that neighborhood residents engage in to control inappropriate public behaviors and thereby prevent conditions which may lead to criminal behavior, has been argued to be both more pervasive and more powerful in controlling crime than formal social control (DeLeon-Granados 1999; Paternoster et al. 1983; Tittle 1980). As Banton (1964:2) noted over 50 years ago, social control is primarily “maintained by the rewards and punishments which are built into every relationship.” While formal law enforcement agencies are certainly important in regulating behavior, the informal social control among neighbors and friends is far more important.

Contemporary versions of social disorganization theory identify neighborhood levels of informal social control as key to understanding variation in crime and delinquency rates across neighborhoods. Within classic interpretations of social disorganization theory, the neighborhood capacity for informal social control was originally viewed to result from the structural characteristics of neighborhoods, including poverty, residential mobility, racial and ethnic heterogeneity, and later, female-headed households. As social disorganization evolved throughout the 1990s, however, it merged with a renewed interest in social networks, drawing extensively from urban sociologists’ systemic model of communities (Kasarda and Janowitz 1974). Within this model of community dynamics, social behavior was not chiefly affected by population size and density, as other urban sociological models suggested, but rather was rooted in “the complex system of friendship and kinship networks and formal and informal associational ties” that existed in the community (Kasarda and Janowitz 1974:330). Incorporating these ideas into the traditional social disorganization model, the systemic model of social disorganization theory argues that neighborhood structural characteristics affect informal social control because they affect the breadth and quantity of social ties within neighborhoods (Bursik and Grasmick 1993).

Knowledge about why social ties are important is often based on ideas about the way connections among residents create various dimensions of social organization particularly social capital (see Coleman 1988; Lee 2008; Putnam 2000; Sampson 2012). The broader concept of social capital, while used in many ways, generally refers to the social relationships and trust among neighborhood residents that allow for the coordination of efforts directed toward mutual benefit (Coleman 1988; Lin 2002; Portes 1998; Putnam 2000). Importantly, according to Lin (2001), it is the actual interactions of residents that is key in the production of social capital. According to this view, it is the “interacting members [of groups] who make the maintenance and reproduction of this social asset possible” (Lin 2001:8). For example, socializing among residents facilitates information sharing, creates an awareness of and support for shared norms, and thereby enhances the collective supervision of children (Coleman 1988; Hagan, MacMillan, and Wheaton 1996; Warner and Clubb 2013).

As studies of the effects of neighborhood social ties on crime and informal social control emerged, however, the findings provided an unclear picture of their implications for informal control capacity. For instance, some studies found that local social ties increased levels of informal social control (Burchfield 2009; Morenoff, Sampson, and Raudenbush 2001; Silver and Miller 2004; Warner 2003), whereas other research found ties to be inconsequential for control efforts (e.g., Carr 2003; Sampson, Morenoff, and Earls 1999), or important only within certain contexts, such as predominantly White neighborhoods (Warner and Wilcox Rountree 1997). Increasingly, social disorganization theorists realized that not all social networks were equally relevant for building informal social control. This latter development generated interest in the effects of different types of networks for the regulatory ability of communities (see Kubrin and Wo 2015; Warner and Clubb 2013). In terms of the patterning of social ties, Bellair’s (1997) research suggests that occasional interactions among residents may be more relevant for informal social control than frequent interactions. And in terms of the types of social ties, Wilcox Rountree and Warner’s (1999) study found that women’s ties to neighbors are important for informal social control, but male’s ties to neighbors are not. Researchers have also found that connections between conventional residents and criminal offenders within communities decrease certain forms of informal social control (Pattillo-McCoy 1999). Social ties in neighborhoods with high numbers of residents enmeshed in criminal activities might not always bring about reductions in crime control efforts.

Within the sociological literature on theories of social organization, specific types of social ties have been discussed in terms of broader concepts such as in-group (“bonding”) and out-group (“bridging”) ties. Putnam (2000, 2007:143) describes bonding ties as “ties to people who are like you in some important way” and bridging ties as “ties to people who are unlike you in some important way.” Bridging ties have been argued to provide a greater breadth of connectivity within a community, and therefore more important for the development of informal social control. Beyerlein and Hipp (2005) examined adherents (and congregations) of different religious traditions, identified as engaging in either predominantly bridging or bonding networks, and found that religious traditions identified as being associated with bridging ties (e.g., mainline Protestants) were related to lower levels of community crime.

Further, closely related to bridging ties are localized ties thought to create intergenerational closure among residents. Here, theorists have argued that intergenerational closure, a social network in which parents know the parents of their children’s friends, is particularly relevant for the regulation and socialization of children within communities (Bursik 1999; Coleman 1988; Sampson et al. 1999). This closure allows parents to communicate about their children’s behavior and develop agreement regarding inappropriate behavior and sanctions. With this shared agreement among the children’s parents, parents are better equipped to monitor the children and the children are aware that sanctions are likely to be applied by their friends’ parents and their own parents. Further, when parents know the parents of their children’s friends, they feel more comfortable and confident disciplining them for rule infractions. This degree of familiarity also provides a basis for a relationship characterized by trust and expectations for action. Building from Coleman’s (1988) insights on this intergenerational closure among children’s parents, some researchers have suggested that these ideas can be generalized to a wider variety of intergenerational ties such as those among children and other community adults. For example, Sampson et al. (1999) note that “[t]he idea of intergenerational closure can also be generalized to include any local adult” (p. 635).

This emphasis on the importance of interactions between youth and responsible adults for a community’s regulation of youth is evident in strands of Shaw and McKay’s (1969) work and the Chicago Area Projects (CAPs) developed by Shaw. In the CAPs, community adults were encouraged to participate in recreational activities with neighborhood youth. In Schlossman and Sedlak’s (1983) review of the CAPs, they note that through interacting with youth, community adults provided important role models for conventional behavior and communicated the importance of prosocial behavior. As Schlossman and Sedlak (1983:428) state, “CAP workers tried to persuade local youth—in a low-keyed, tolerant, non-abusive manner—why it was both morally right and, ultimately, in their best interest to conform to the values and expectations of conventional society.” The CAPs reliance on adult–youth interactions in day-to-day social activities was believed also to be important because over time these interactions would “build up a sense of neighborhood responsibility for monitoring youth conduct” (Schlossman and Sedlak 1983:403, emphasis added). More recently, informed by the CAPs conceptual framework, researchers implemented an intervention program in low-income Atlanta areas whose purpose was to forge strong relationships between pairs of adults (65 percent male) and area children who were recruited for the study and then to provide them with various mentoring skills for the purpose of fostering enhanced informal social control (Ohmer 2016). Participants reported increased engagement and connections in their communities after the intervention as well as heightened levels of trusting social relations with their neighbors (p. 700).

Recent qualitative work has also pointed to residents’ acknowledgment of the importance of responsible adults who ensure that children follow the rules and are safe, often lamenting the loss of adults willing to fulfill those roles (Cantora, Iyer, and Restivo 2016; De-Leon Granados 1999). For example, in Cantora et al.’s (2016) qualitative study in East Baltimore that centered on identifying neighborhood problems and solutions, one male resident noted, “Everyone is defending by themselves, it used to be that if one parent was outside, they would watch over the other kids. People would volunteer to be responsible” (p. 703).

In the more general sociological literature, dense relational networks involving children and adults in communities have been viewed as important as they “transcend the household,” enabling a larger number of adults to observe and discuss the behaviors of area children in different circumstances (Wilson 1996:62). As noted above, where adults and children engage in frequent social exchange, adults may be more likely to take a collective responsibility for children’s behavior.

The continuing interest in the implications of different types of ties for generating community controls coupled with questions about the positive roles adult males might have in the broader community reveals an overlooked but important organizational feature of communities: the intergenerational ties between adult men and neighborhood youth. We see these intergenerational social ties as a potentially important way in which male ties may also matter for the informal social control of youth.

Current Study

Research within social disorganization theory’s systemic model continues to probe the dynamics of local network structures to understand how communities generate informal social control. According to contemporary theoretical models, the absence of males within families is generally considered an important and common feature of high crime communities. While this deficit model accentuates the important role of men within families, there is limited empirical evidence regarding how males in the broader community may also add to the social organization of the community to prevent crime. Drawing on the social disorganization and urban underclass literatures, this study tests whether male ties with local youth within neighborhoods are important contextual determinants of the willingness to engage in informal social control. Specifically, does the volume of social ties between adult males and youth provide a unique positive effect on the perceived likelihood of informal social control of children separate from the simple presence of males and other measures of social ties between adults?

Measures

Dependent variables

The main dependent variable in this study is a measure of the likelihood of informal social control specifically directed at children. It is a measure of the perceived likelihood of residents intervening with children involved in inappropriate behavior. Specifically, the measure was derived from the following five questions: “If a teenager from your neighborhood was seen away from the school grounds on a school day, how likely is it that someone in your neighborhood would ask that child why he or she wasn’t in school?” “If elementary school age children were on the street, in a park, or in a store on a school day, how likely is it that someone in your neighborhood would ask them why he or she wasn’t in school?” “How likely is it that someone in your neighborhood would tell the child’s parents?” “If children were spray painting on a local building, how likely is it that someone in your neighborhood would do something to stop it?” and “If children were showing disrespect to an adult in your neighborhood, how likely is it that someone else in your neighborhood would do something to stop it?” These items follow previous research on communities examining similar informal social control behaviors (see, e.g., Elliott et al. 1996; Sampson, Raudenbush, and Earls 1997). Response categories for each of the questions were “very likely,” “somewhat likely,” “somewhat unlikely,” and “very unlikely,” with responses (re)coded such that higher scores indicate higher likelihoods of intervening. The items were averaged for all respondents that answered at least three of the five items to create an index of informal social control of children. ⁴ The α for these items was .816.

Because the importance of informal social control is due to its effect on crime rates, our final model examines the effects of informal social control on the part I crime rate for the year 2000. The addresses of reported crimes were obtained from the police departments of both cities and were geocoded into each neighborhood using ARCGIS software. Neighborhood counts of part I crime incidents for 2000 were divided by the 2,000 population count for each neighborhood and multiplied by 1,000 to get the total reported crime rate per 1,000 population. We use the total number of reported crimes for each neighborhood as higher levels of informal social control are expected to decrease the overall level of crime. Note, however, it was necessary to exclude aggravated assaults from our total measure of crime because assaults were classified somewhat differently by police officials across the two cities. ⁵

Independent variables

Our key independent variable is the neighborhood level of adult male–youth social interactions. It assesses the extent to which adult males are visibly involved with children in the neighborhood. This measure of social interaction among men and youth captures a form of social ties that is not generally measured in the social disorganization literature and may capture an important way in which male ties matter for informal social control. Our measure of this concept is based on the question, “How common is it to see adult men in this neighborhood engaged in play or educational activities with neighborhood children?” The response categories for this question were “very common,” “somewhat common,” “somewhat uncommon,” and “very uncommon.” We assess the proportion of residents within each neighborhood who responded it was “very common.” We believe that while responses of “very common” (vs. “somewhat common”) provide a more conservative estimate of actual male–youth social interaction, it is likely to be more valid than a less polar response. As noted previously, a preliminary examination found that in 2 of the 66 neighborhoods 30 percent or more of the respondents stated that there were no children in the neighborhood. Because of the high percentage of respondents not providing a valid response on this variable, we deleted these two neighborhoods from the analyses (n = 64).

In assessing the effects of male–youth interactions on the willingness to engage in informal social control, it is important to control for other types of social ties to be certain that this measure is not simply carrying the shared variance with other forms of ties that are related to informal social control. Measures representing other forms of social ties within the neighborhood include neighboring social ties, friends and family ties, ties to neighborhood organizations, and social cohesion and trust among neighbors. The neighboring social ties variable is measured as the average number of neighboring activities engaged in at least monthly within each neighborhood. Specifically, respondents were asked how often they engaged in six activities with neighbors (having a neighbor over to your house or going to their house, borrowing or exchanging items with neighbors, asking for help, going out for an evening with a neighbor, talking about sales or programs for kids, and talking about personal problems). Response categories were “about once a day,” “about once a week,” “about once a month,” “several times a year,” “about once a year,” and “never.” Responses of “about once a day,” “about once a week,” and “about once a month” were recoded to 1 and responses of “several times a year,” “about once a year,” and “never” were recoded to 0. Responses to the six recoded items were then summed across all the neighboring activities, providing a measure of neighboring activity engaged in regularly. These sums were then averaged across all individuals within a neighborhood to create the neighborhood-level index of average neighboring social ties. The Cronbach’s α for this index was .79.

We also measure friends and family-based social networks within neighborhoods with a separate set of survey items. Respondents were asked, “How many of your relatives live in your neighborhood, not including those in your household?” and “Not including people in your household, how many of your neighbors do you consider to be friends?” Because of some extreme outliers on both items, the number of relatives in the neighborhood was truncated to 20 (which included 99.6 percent of the actual responses) and the number of neighbors considered friends was truncated at 50 (which included 99.3 percent of the actual responses). All values greater than these were recoded to 20 and 50, respectively. Responses to both items were summed for each respondent and then averaged across all respondents in the neighborhood to obtain the neighborhood-level measure of friends and relatives ties. For a portion of the analysis, we disaggregate each of the foregoing social ties measures by the gender of the respondent to provide the average number of gender-specific neighboring ties and familial and friendship ties for each neighborhood.

Social cohesion and trust is also an important aspect of connectivity within neighborhoods and is an important social organizational foundation for effective informal social control within communities (Sampson 2012). It is measured with the following four items: (1) “people in my neighborhood care about the neighborhood,” (2) “people in my neighborhood can be trusted,” (3) “people in my neighborhood share the same values,” and (4) “people in my neighborhood generally get along with each other.” Responses to these questions were “strongly agree,” “somewhat agree,” “somewhat disagree,” and “strongly disagree.” Responses were (re)coded such that higher scores reflect stronger levels of agreement. Responses to each of the items were then averaged as long as the respondent answered at least three of the four items. These averages were then aggregated to the neighborhood level. Cronbach’s α for this index is .78. ⁶

We also include a measure of social ties based on organizational participation in informal social organizations in each neighborhood as an indicator of the average exposure of residents to organizational networks and their involvement in organizations conducive to informal control (Bursik and Grasmick 1993). The measure captures the proportion of respondents in each neighborhood who responded “yes” to the question, “Do you attend meetings of any neighborhood-based groups or organizations, such as neighborhood churches, parent-teacher associations (PTAs), neighborhood associations, watch group, political associations, and so on.” The responses were “yes” or “no.”

Because social disorganization theory argues that each of these measures of connectivity within neighborhoods is caused by neighborhood structural conditions, we also include in our models measures of neighborhood disadvantage and residential mobility. Measures of disadvantage and mobility are based on 2000 census data at the block group level. Seven items frequently used to measure disadvantage and mobility from the census were factor analyzed and two factors with eigenvalues greater than one emerged. The items and their factor loadings for the disadvantage factor were percent on public assistance (.87), percent African American (.87), percent female-headed households with children (.78), percent with less than a high school degree (.74), and percent below poverty (.74). Percent renters (.87) and percent of residents that lived in a different house five years prior (.94) loaded on the mobility factor. Regression-based factor scores were computed for both factors.

To account for the visibility of residential public activity across neighborhoods, or what we consider residents who spend time outside, we include scores from a measure that asks respondents, “How common it is for people in your neighborhood to spend time outside of their home for more than just a few minutes, sitting outside, working in the yard, or taking a walk in the neighborhood?” Response categories were “very common,” “somewhat common,” “somewhat uncommon,” and “uncommon.” For each neighborhood, we calculate the proportion of respondents who report it is “very common” to see residents spend time outside of their home. Finally, because the extent to which men engage in activities with children and the perceptions of informal social control may both be affected by the level of men in the neighborhood, we control for the proportion of the neighborhood population that is male (proportion male) using census data on the number of males and the total neighborhood population. This measure also allows us to distinguish between the presence of males in general and male interactions with youth. We also include a dummy variable for the city where the neighborhood is located to account for potential unmeasured differences between the two surveyed cities that could affect the hypothesized associations.

It is important to control for crime rates in the multilevel regression models as well, since any effect of social organizational resources on informal control might be confounded by the past volume of crime in a neighborhood. We control for a lagged measure of neighborhood crime rates using official reports of crime. Counts of different types of reported crimes within neighborhoods were calculated from the geocoded police data mentioned above. We developed a lagged crime rate measure based on the summed count of crime for 1997 and 1998 using the following crimes: homicide, commercial robbery, personal robbery, residential burglary, and commercial burglary. We focus on these serious crimes as these are the types of crimes that are most likely to decrease community interaction and subsequently informal social control. This count was divided by two to obtain the average annual number of crimes for each neighborhood. We then calculated a crime rate by dividing this average annual crime count by the 1990 neighborhood census population and multiplying that value by 1,000. We selected these years because they are before the year in which the survey was administered to the residents. As noted above, we do not include aggravated assault in this measure because of the differences between the two cities in assault classifications.

Neighborhood effects can be either contextual or compositional, that is, they can be due to emergent characteristics of the neighborhood or due solely to the differences in the characteristics of residents that comprise the neighborhood. Therefore, the models examined here control for relevant individual background characteristics. Specifically, the following individual-level measures variables from the survey were included age (in years), gender (1 = male), race (1 = African American), marital status (1 = single), education (1 = less than a high school degree), and the number of children under the age of 18 in the respondent’s household. Table 1 displays the descriptive statistics for the array of study variables.

OPEN IN VIEWER

Table 1.

Descriptive Statistics for Study Variables.

Study Variables	Mean	SD	Minimum	Maximum	N
Dependent variables
Informal social control	2.80	0.89	1.00	4.00	2,157
2000 Crime rate per 1,000 population	93.54	77.54	3.50	360.66	64
Neighborhood level
Disadvantage	0.00	1.00	−2.19	2.17	64
Mobility	0.00	1.00	−2.18	1.76	64
Lagged crime rate (1997/1998) per 1,000 population	28.52	24.32	0.79	109.15	64
Proportion residents outside	0.72	0.11	0.35	0.92	64
Proportion organizational participation	0.39	0.13	0.14	0.77	64
Social cohesion and trust	3.17	0.36	2.40	3.93	64
Female neighboring social ties	2.21	0.47	0.90	3.15	64
Male neighboring social ties	2.33	0.77	1.00	4.40	64
Female friends–relatives social ties	7.10	2.55	2.59	15.48	64
Male friends–relatives social ties	8.84	3.88	3.36	21.22	64
Friends–relatives social ties	7.54	2.44	3.66	16.83	64
Neighboring social ties	2.24	0.42	1.17	3.26	64
Adult male–youth interactions	0.16	0.08	0.00	0.43	64
Proportion males	0.48	0.07	0.35	0.69	64
City of residence	1.50	0.50	1.00	2.00	64
Individual level
Less than high school	0.20	0.40	0.00	1.00	2,236
Age	46.32	16.88	18.00	94.00	2,228
African American	0.50	0.50	0.00	1.00	2,221
Male	0.32	0.46	0.00	1.00	2,238
Single	0.34	0.47	0.00	1.00	2,232
Number of children	0.67	1.12	0.00	9.00	2,239

Results

Table 1 presents descriptive data on all the variables in the models. As can be seen in this table, the average perceived likelihood of informally controlling children is 2.80, suggesting that the average likelihood is close to “somewhat likely,” but it varies across these neighborhoods from 1 (very unlikely) to 4 (very likely). Similarly, as can be seen in Table 1, measures of social organization at the neighborhood level vary across neighborhoods. The average score of neighboring ties is 2.24 with a range of 1.17–3.26 across the 64 neighborhoods, with males reporting slightly higher levels of neighboring than females (2.33 vs. 2.21). The average number of ties to friends and family is 7.54 across neighborhoods, but this varies between 3.66 and 16.83. Again, in terms of quantity of ties to friends and family, males tend to have a slightly higher mean (8.84) than the females (7.10). The average level of social cohesion and trust across neighborhoods is relatively high, 3.17, but varies from 2.40 to 3.93. In terms of adult male–youth interactions, the average neighborhood proportion of respondents stating it was “very common” to see men playing with children was somewhat low at .16. This proportion varied across neighborhoods from a low of .00 to a high of .43. The average proportion of residents that are males in these neighborhoods was .48, with the range being from .35 to .69. The average proportion of respondents in each neighborhood that participated in neighborhood organizations was .39, ranging from .14 to .77. The average proportion of residents stating it was very common to see residents spending time outside was .72, suggesting that residents should be well aware of what is going on in their neighborhoods. The average lagged crime rate for the five crimes included in this measure across 1997 and 1998 was 28.52 per 1,000 residents, and the average total crime rate for 2000 (without assault) was 93.54 per 1,000 residents. Neighborhoods were evenly distributed across both cities.

In terms of the characteristics of the sample, 34 percent were single, 50 percent were African American, and 32 percent were male. Twenty percent of respondents did not have a high school degree, and the average number of children living in the home was less than one, ranging from 0 to 9. The average age of respondents was 46 years old.

Before beginning our multivariate analyses, we examined the null model (the equation without any of the predictor or control variables) to determine whether the mean level of informal social control of children varied significantly across neighborhoods. Results from this model (not shown) demonstrate that the intercept for informal social control of children does vary significantly across neighborhoods as the variance component (u_0j) is significantly greater than zero (χ² = 135.089, df = 63, p = .000). However, as with much neighborhood-level research, the variation between neighborhoods accounts for only a small amount of the total variance, approximately 3.1 percent.

We next ground our analyses in the basis of our argument, derived from past research, that male ties have not been found to make a significant contribution to informal social control. First, following Wilcox Rountree and Warner (1999) who used data from Seattle, we examine the effects of gender-specific neighboring social ties on informal social control. We then expand to also examine another common measure of social ties among neighbors—friends and relatives ties—by gender. Each of these initial models control for individual-level measures of respondent personal characteristics and traditional measures of neighborhood structure (disadvantage and mobility), the lagged crime rate, social cohesion and trust, organizational participation, and the proportion of residents that spend time outside. These two models provide information about the role of male versus female social ties across different conceptualizations of social ties. As can been seen in both models (models 1 and 2 in Table 2), female ties significantly increase informal social ties, while male ties do not. Specifically, in model 1, the results indicate that female neighboring social ties (b = .123, SE =. 042) significantly increased informal social control of children, whereas the effects of male neighboring social ties are not significant determinants of informal control (b = .011, SE = .024). In model 2, female ties to friends and relatives also significantly increased informal social control (b = .025, SE = .007), while male ties to friends and relatives had no significant effect (b = .006, SE = .004). In both models, neighborhood levels of social cohesion and trust significantly increased informal social control, as did the number of children in the household at the individual level. Being male significantly decreased the perception of the likelihood of informal social control. In model 1, examining neighboring social ties, respondents that are African American also reported higher levels of informal social control. Note that a preliminary model (not shown) examining only the structural characteristics of the neighborhood—disadvantage, residential mobility, and the lagged crime rate—did find that mobility also significantly decreased informal social control (b = −.133, SE = .030). As predicted by social disorganization theory, this effect is mediated in the models that account for neighborhood measures of social ties. ⁷

OPEN IN VIEWER

Table 2.

Multilevel Linear Regression Models Predicting Informal Social Control.

	Model 1	Model 2	Model 3
	Coefficient (SE)	Coefficient (SE)	Coefficient (SE)
Neighborhood level
Disadvantage	−0.010 (0.041)	0.007 (0.042)	−0.016 (0.041)
Residential mobility	−0.053 (0.037)	−0.021 (0.037)	−0.027 (0.034)
Lagged crime rate (97–98)	0.0003 (0.0008)	0.0002 (0.0007)	0.0002 (0.0008)
Residents outside	0.314 (0.225)	0.189 (0.239)	0.181 (0.225)
Organizational participation	0.265 (0.225)	0.252 (0.149)	0.199 (0.136)
Social cohesion trust	0.332** (0.108)	0.271** (0.109)	0.271** (0.100)
Female neighboring social ties	0.123** (0.042)
Male neighboring social ties	0.011 (0.024)
Female friends–relatives ties		0.025** (0.007)
Male friends–relatives ties		0.006 (0.004)
Neighboring social ties			0.082 (0.043)
Friends–relatives ties			0.026** (0.006)
Adult male-youth interactions			0.361* (0.171)
Proportion males			−0.185 (0.433)
City of residence	0.012 (0.041)	0.041 (0.041)	0.036 (0.039)
Individual level
Less than high school education	0.033 (0.071)	0.025 (0.071)	0.027 (0.071)
Age	−0.001 (0.001)	−0.001 (0.001)	−0.000 (0.001)
African American	0.124* (0.061)	0.120 (0.061)	0.127* (0.061)
Male	−0.088* (0.042)	−0.087* (0.042)	−0.087* (0.041)
Single	−0.047 (0.046)	−0.049 (0.046)	−0.047 (0.045)
Number of children	0.066** (0.018)	0.067** (0.018)	0.066** (0.018)
Constant	1.048** (0.358)	1.372** (0.345)	1.269** (0.434)
Respondents	2,127	2,127	2,127
Neighborhoods	64	64	64

*p < .05; **p < .01.

Having verified that, consistent with our argument, traditional measures of male social ties do not appear to matter for the willingness to engage in informal social control, we next turn to our main analysis. In this model (model 3), we examine a specific type of male tie that has not previously been empirically analyzed—adult male–youth interaction. Because we have already demonstrated that male ties do not have an effect as traditionally measured, and in order to decrease multicollinearity and keep the number of neighborhood variables at a reasonable amount, we replace the gender-specific measures of both neighboring ties and friends and relatives ties with their respective overall or omnibus neighborhood measures. We also add a measure controlling for the proportion of males in the neighborhood. The latter measure is needed to account for the possibility that some of the adult male–youth interactions measure may be conflated with the simple presence of males in neighborhoods where they represent a larger fraction of the area population.

Findings from this equation (model 3) show that observed adult male–child interactions have a significant and positive effect (b = .361, SE = .171) on informal social control that is unique and in addition to the effects of the traditional measures of neighboring ties (b = .082, SE = .043) and ties to friends and relatives (b = .026, SE = .006). The results also show that the proportion of males in the neighborhood, in and of themselves does not significantly affect the perceived likelihood of informal social control. These findings suggest that, beyond the levels of social cohesion and trust and social ties among adult neighbors, the extent to which adult males have relationships with neighborhood youth increases residents’ perceived likelihood of informal social control. Furthermore, the effects of social ties to friends and family, and social cohesion and trust remain significant determinants of informal control with the addition of the male ties to youth variable, suggesting each is a unique form of social organization of importance to the social control of children.

Finally, because the importance of informal social control relies heavily on its theorized impact on neighborhood crime levels and because we use a rather unique measure of informal social control (informal social control of children), we provide a model in which we examine the effects of our informal social control measure on the overall neighborhood crime rate, controlling for the other variables in the model. In this analysis, our dependent variable is now a neighborhood-level variable; therefore, we analyze the model with ordinary least squares regression using only our neighborhood-level predictors. As can be seen in Table 3, informal social control significantly decreases the neighborhood crime rate. Additionally, the lagged crime rate is significantly and positively related to the current year crime rate and the proportion males in the neighborhood is also associated positively with the crime rate. None of the measures of neighborhood connectivity were found to be directly related to crime, but rather as theorized, these variables are associated with neighborhood levels of informal social control which is the key intervening variable between neighborhood social ties and crime. ⁸

OPEN IN VIEWER

Table 3.

Linear Regression Model Predicting Year 2000 Neighborhood Crime Rate.

Independent Variables (N = 64)	Coefficient	SE
Disadvantage	3.777	12.981
Residential mobility	5.024	12.826
Lagged crime rate (97–98)	1.182**	0.348
Residents outside	−49.211	87.231
Organizational participation	55.209	67.409
Social cohesion and trust	18.152	43.949
Neighboring social ties	38.723	23.322
Friends–relatives social ties	4.577	4.257
Adult male-youth interactions	−91.900	101.074
Proportion males	328.333*	140.407
City of residence	−15.073	16.864
Informal social control	−127.557*	54.506
Constant	129.900	167.3454
Adjusted R2	.449

Note: N = 64.

*p < .05; ** p < .01.

Discussion and Conclusion

Previous research within social disorganization theory has demonstrated how the presence of female-headed households within communities weakens the capacity of communities to regulate crime and other inappropriate public behavior within their borders. Those findings suggest the absence of adult males in families creates a collective deficit in resources for informal social control. Yet, aside from the study of men’s absence within the home, few studies have specifically examined how, or if, men might add to the social capital available in the broader community. This is particularly curious given recent advances in the systemic model of social disorganization theory, which have highlighted the centrality of social ties and social networks in creating social capital within communities, a key resource for the development of informal social control. Moreover, social disorganization theories have increasingly recognized that not all ties are equal in terms of their resource potential; different types of ties and those in different contexts have been found to vary in their contributions to informal social control.

In the current study, we have broadened investigations of social ties to include male intergenerational ties to community youth. While we have added to previous findings that demonstrate male ties as traditionally measured in social disorganization studies (neighboring activities and number of friends) do not increase the willingness to engage in informal control, we have also found that a different type of male social tie is important for assessing men’s contributions to neighborhood social organization. Specifically, findings from this study show that adult males’ ties with youth increase the perceived likelihood of informal social control of children within the community. Importantly, these effects are in addition to the effects of two traditional measure of social ties (neighboring activities and numbers of friends and family ties), a measure of social cohesion and trust, and controls for the presence of males the neighborhood. This finding suggests that male intergenerational ties with youth have their own specific and important influence on a community’s capacity for crime control. While men’s neighboring activities and the size of their neighborhood networks are found to be less important than women’s, men, nonetheless, contribute importantly to the overall social capital of the community for crime control.

What is it then that makes this intergenerational connection between adult males and youth important for informal social control when male ties with neighbors and neighborhood friends are not important? It is likely that male ties to neighbors, while more frequent and numerous than female ties (in these data) provide a different form of social capital. Social networks become social capital when they perform a function that helps to achieve a goal for the collective, such as informal social control. In most of the social disorganization literature, social networks are viewed as creating social capital because they function to increase trust and social cohesion among neighbors. Social cohesion and trust is an important basis for informal social control because it produces the obligation among neighbors to support one another in their efforts to provide social control. However, male social networks with adult neighbors appear to be less capable of building shared trust and social cohesion than female networks. The correlation between our measure of females’ friends and relatives is significantly correlated with our social cohesion and trust measure (r = .403, p < .01), while the measure of males’ friends and relatives is not correlated with social cohesion and trust (r = .149). Intergenerational ties (measured here in terms of adult male ties to youth) may work through a different mechanism, as our measure of adult male interactions with youth are also not correlated with social cohesion and trust (r = .045). These ties are therefore likely to provide a different type of social capital. Through engaging with youth in prosocial activities, adult male residents may provide a form of social capital that strengthens the norm of acting in the best interest of the collectivity as well as strengthening appropriate behavioral norms (see Coleman 1988).

As noted above, this intergenerational activity (of males) may remind other residents of broader obligations and expectations for interacting with and regulating youth, challenging the laissez-faire “these kids are not my kids” mentality. Relatedly, these visible intergenerational activities in the neighborhood may signal that these men are socializing youth into conventional values, which may in turn make others’ willingness to sanction youth more likely, or perhaps, contact those adult males to address problems with youth when they arise. With these data, we are not able to determine what explicit form of social capital arises from these male intergenerational ties to youth, but it does appear to be different from the form of social capital that arises from female ties with neighbors.

Of course, we are attuned to the potential scholarly criticisms surrounding the suggestion that male social ties with youth create positive effects for the structure and nature of neighborhood social organization as it relates to crime control resources. In some neighborhoods, one might expect that such ties could involve negative influences that would further undermine community safety. Indeed, we do not suggest that all such ties are positive in terms of how they affect youth behavior, only that adult male ties to youth provide increased informal social control of youth. Clearly, the effects of social ties within communities are complex and nuanced. Nonetheless, our findings lead us to side with Rose and Clear’s (1998) suggestion that even males who have committed crimes and return to their communities after incarceration can still provide positive social resources to their communities. Further, our findings are in line with the ideas of Shaw and McKay (1969) as they were implemented in the CAPs projects, which emphasized the importance of intergenerational ties within the community through recreational and social activities (see Schlossman and Sedlak 1983; Warner, Beck, and Ohmer 2010).

Our findings add to the literature on social ties and informal social control in two important ways. First, they draw attention to the understudied area of the potential positive role of male intergenerational ties within the broader community. While there is no doubt that social ties to adult males can sometimes dampen informal social control resources, such as in the case of ties to criminal gangs (e.g., Pattillo-McCoy 1999), and organized crime because they provide a form of criminal capital, there has been little criminological research that has attempted to examine their positive imprint on the broader community, particularly in relation to informal social control efforts. Second, our findings add to recent interest on the importance of the breadth or nature of social ties for informal social control and collective behavior (Bursik and Grasmick 1993; Putnam 2007). Several recent studies have pointed to bridging ties as an important dimension of connectivity within the community because they facilitate the diffusion of information, resources, norms, and expectations across groups (Putnam 2000). Previous studies have found that bridging ties in terms of ethnicity (Warner, Swartz, and Hawk 2015) and religious groups (Beyerlein and Hipp 2005; Warner and Konkel 2019) increase informal social control or decrease crime. The visibility of ties between adult men and youth appears to be an additional bridge across groups that adds to the neighborhood’s likelihood to provide informal social control.

Although we believe this study opens an important area of research in social disorganization theory, the study is not without limitations. First, we were not able to assess information on frequency or type of activities that adult males engage in with youth. While we believe such data would have been informative, we also believe that resident’s perceptions of the role of men with youth may be equally important. That is, the collective perceptions of whether men are actively involved in the lives of local youth might be most critical for whether area residents are willing to engage in informal control. Second, we also were unable to ascertain whether the adults seen interacting with community youth are doing so with their own children or community children in general. The survey question simply is not constructed to capture such level of detail. Nonetheless, it may be that some residents observing these interactions would also not know whether the youth were family members or other community members, and it may simply be the presence of these interactions that matters most for the perceived willingness to engage in social control. Third, our measure of male ties to youth is based on only one indicator from survey data. A more reliable measure would be based on an index of several items related to male activity with youth. Finally, because the data are cross-sectional, we are not able to determine whether prior levels of informal social control influenced the frequency of interactions between adult males in the community and children.

Finally, we note that our arguments and findings regarding how intergenerational ties may lead to increased informal social control of youth are not necessarily gendered. Female ties to youth may fulfill similar roles. Our reasoned intent was simply to determine whether male social ties examined as intergenerational ties may be a way in which male ties also matter. Our findings suggest they do, even in the presence of both gendered and nongendered traditional measures of social ties. We cannot, however, with the data at hand, claim that male ties with youth are more important than females’ ties with youth. We can only speak to the importance of adult male-to-youth social interactions and suggest it would be fruitful for criminologists to examine the contributions of female-to-youth ties as well. Finally, our study does not address whether male intergenerational interactions would have different effects based on the gender of the children. We would expect, though, that adult male youth interaction would be consequential for perceived informal social control regardless of the gender of children. These possibilities await future testing.

Although we believe that our findings add to emerging understandings about the complexity of the effects of social ties on communities’ capacities for crime control, many questions remain. Among these are questions about how the nature of different ties may work in tandem with or counter to each other. Are women’s friendship ties enhanced by the presence of male ties with youth? Do male intergenerational ties with youth counterbalance the diminished effects of female ties in neighborhoods with higher rates of female-headed households? Do the positive effects of male interpersonal ties on the willingness of individual to engage in informal social control vary across community contexts? Indeed, there is still much to be learned about the complex and rich tapestry of community ties and how they support or inhibit community crime control. Part of this effort we believe should include further exploration of the potential roles of men for informal crime control efforts, not by their absence, but through the roles that they may specifically, or perhaps, uniquely contribute to community social organization.