writing assignment

i m Poverty? Race, o f L o w - S k i l l e d

gers at the Gates: irica, edited by R. ;rsity of California

ichter. 2003. How ion and the Social y, C A : University

Society. Berkeley, :ess. eroskedasticity- Estimator and a

ry." Econometrica

ihood Estimation metrica 50:1-25. i Disadvantaged: md Public Policy. ago Press. pears: The World York: Alfred A .

M . Neckerman. Structure: The and Public Policy ? Poverty: What 1S. Danziger and irvard University

md Moral Order. fornia Press, inants of Recent ;." International

innarelli. 2001. 'are Programs: deralism: Issues igton, DC: The

Loprest. 2001. Disadvantaged New World of

i d R. Haskins. Institution. 98. Growing up •en Adapt to Life : Russell Sage

Violence, Older Peers, and the Socialization of Adolescent Boys in Disadvantaged Neighborhoods

David J. Harding University of Michigan

Most theoretical perspectives on neighborhood effects on youth assume that neighborhood context serves as a source of socialization. The exact sources and processes underlying adolescent socialization in disadvantaged neighborhoods, however, are largely unspecified and unelaborated. This article proposes that cross-cohort socialization by older neighborhood peers is one source of socialization for adolescent boys. Data from the National Educational Longitudinal Survey suggest that adolescents in disadvantaged neighborhoods are more likely to spend time with older individuals. I analyze qualitative interview data from 60 adolescent boys in three neighborhoods in Boston to understand the causes and consequences of these interactions and relationships. Some of the strategies these adolescents employ to cope with violence in disadvantaged neighborhoods promote interaction with older peers, particularly those who are most disadvantaged. Furthermore, such interactions can expose adolescents to local, unconventional, or alternative cultural models.

Most theoretical perspectives on neighbor-hood effects on youth assume that the neighborhood serves as a source of socialization, particularly for adolescents. Through differen- tial exposure to behavioral models or cultural ideas, disadvantaged neighborhoods are thought to influence how young people make decisions in domains such as schooling and romantic rela-

tionships. For example, Wilson's (1996) social isolation theory argues that residents o f poor neighborhoods are isolated from middle class or mainstream social groups, organizations, and institutions as a result of joblessness. Social isolation creates cultural isolation, which— when combined with diminished educational and labor market opportunities—leads to the

Direct correspondence to David J. Harding at Department of Sociology, University of Michigan, ~ 500 S. State St., A n n A r b o r , M I 48109-1382 ([email protected]). Funding for this research was provided by the National Science Foundation (SES-0326727), The William T. Grant Foundation, the American Educational Research Association/Institute o f Education Sciences, the MacArthur Foundation Network on Inequality and Economic Performance, and the Harvard M u l t i d i s c i p l i n a r y Program on Inequality and Social Policy, which is funded by an NSF Integrative Graduate Education and Research Traineeship grant. A n N I C H D Post-Doctoral Fellowship at the Population Studies Center at the University o f Michigan provided additional support. Katherine Newman, Christopher Winship, Robert Sampson, Christopher Jencks, Jeff Morenoff, A l

Young Jr., Renee Anspach, Andrew Clarkwest, Brian Goesling, the ASR editors and reviewers, and audi- ences at the C U N Y Graduate Center, Temple University, Harvard University, UCLA, UC-Berkeley, University o f Michigan, University o f Wisconsin, Cornell University, New York University, and the University o f Chicago provided helpful comments on previous versions o f this article. Shutsu Chai, Johnathan Smith, and Stephen Rose provided excel- lent research assistance during the fieldwork, and Shutsu Chai, Stephen Rose, Kai Jenkins, Lauren Galarza, Aghogho Edevbie, Meaghan Cotter, Amanda Braun, and Randall Monger worked tirelessly to code the data. I thank Anthony Braga of Harvard University and Carl Walter o f the Boston Police Department's Boston Regional Intelligence Center for providing the incident data presented in Table 2.

AMERICAN SOCIOLOGICAL REVIEW, 2 0 0 9 , V O L . 74 (June:445-464)

4 4 6 A M E R I C A N S O C I O L O G I C A L R E V I E W

development o f cultural repertoires that differ from those of working- or middle-class com- munities. Neighborhood interactions can thus expose young people to "ghetto-specific" cul- tural models.

The exact sources and processes underlying such adolescent socialization are, however, largely unspecified. Wilson (1996) argues that in neighborhoods with high joblessness, life is not organized around work, and the experience of observing adults going to work or discussing their jobs is absent. Yet critics counter that even in high joblessness neighborhoods, many peo- ple are working or going to school and that neighbors can just as easily serve as models of what not to do (Newman 1999). While many African American middle-class families leave ghetto neighborhoods, many also remain (Patillo-McCoy 1999). Moreover, both survey and ethnographic research show strong support for mainstream ideals regarding work, educa- tion, marriage, and cmldbearing among the poor (Edin and Kefalas 2005; Goldenberg et al. 2001; Newman 1999; Solorzano 1992; Young 2004). What then are the sources of alternative or ghet- to-specific cultural models in poor neighbor- hoods? Who exposes adolescents to these ideas, and why do adolescents take these individuals seriously, often against the wishes of parents, ministers, teachers, and other community adults?

This article proposes that at least part o f the answer lies in cross-cohort socialization. This is the socialization of adolescent boys by older neighborhood peers, particularly those most likely to be available in the neighborhood: older men who are unemployed, not in school, and involved in the underground economy. Drawing on representative survey data from the National Educational Longitudinal Survey (NELS), I show that, compared with their counterparts in more advantaged neighborhoods, adolescents in disadvantaged neighborhoods are more likely to spend time with older peers. The majority of this article attempts to understand the sources and implications of this pattern. Based on in-depth, unstructured interviews with 60 adolescent boys and their parents in three Boston communities, I describe the role older peers play in the social- ization o f adolescent boys in two poor areas. Violence in these areas reinforces the neigh- borhood as a form of social identity, restricts adolescent boys' pool of potential friends, struc-

tures their use of geographic space, and leads younger adolescent boys to greater interaction with older adolescents and young adults on the street. A l l of these are unintended consequences of boys' strategies for navigating dangerous streets. Older males, particularly those who are unemployed and out o f school, become an important potential source of neighborhood socialization for these boys and have the power to influence their decision making in domains beyond safety. By contrast, adolescent boys in a third Boston community with a low poverty rate and less violence have few older friends and acquaintances and experience less exposure to such socialization.

N E I G H B O R H O O D EFFECTS A N D A D O L E S C E N T S O C I A L I Z A T I O N

Social isolation theory argues that lack of par- ticipation in the mainstream labor market iso- lates residents of inner-city communities from middle-class social groups, organizations, and institutions (Wilson 1996). According to this theory, children in high joblessness neighbor- hoods do not experience life organized around work, and illicit income reduces attachment to the labor market. Social interaction in isolated neighborhoods leads to the development of cul- tural repertoires that differ from the mainstream. Socially and culturally isolated from wider soci- ety and without resources or opportunities for social mobility, residents' social problems spi- ral into a vicious cycle.

Social isolation theory assumes that neigh- borhood interactions expose young people in poor neighborhoods to neighborhood-specific cultural models that differ from those common among the middle class.1 Yet there is strong support for conventional ideals about educa- tion, work, welfare, and marriage among the poor (Edin and Kefalas 2005; Goldenberg et al. 2001; Newman 1999; Solorzano 1992; Young 2004). It is unclear then, who transmits ghetto-

1 Quinn and Holland (1987:4) define "cultural models" as "presupposed, taken-for-granted models o f the world that are widely shared (although not necessarily to the exclusion of other, alternative mod- els) by the members of a society and that play an enormous role in their understanding o f that world and their behavior in it."

V I O L E N C E , A D O L E S C E N T B O Y S , A N D O L D E R P E E R S 4 4 7

e, and leads r interaction idults on the Dnsequences ; dangerous lose who are become an ighborhood /e the power ; in domains cent boys i n low poverty r friends and exposure to

O N

lack of par- market iso- unities from zations, and ding to this ;s neighbor- ized around tachment to i in isolated ment of cul- mainstream. 1 wider soci- rtunities for oblems spi-

that neigh- g people in )od-specific ise common re is strong oout educa- among the

enberg et al. 992; Young mits ghetto-

ine "cultural anted models although not ;mative mod- that play an

of that world

specific or unconventional cultural models to young people in disadvantaged neighborhoods and how these models pass from generation to generation. A purely structural account, in which each succeeding generation faces similar struc- tural barriers and develops similar cultural adap- tations, does not suffice. Because adolescents respond differently to structural limitations, as evidenced by considerable variation w i t h i n neighborhoods in outcomes, response to blocked opportunities cannot be the whole story. Even in the poorest neighborhoods, dropping out of high school, joblessness, and teenage pregnan- cy are far from universal.

One potential mechanism for these cultural transmission processes that has not been fully investigated is peers. As youths enter adoles- cence, the focus of their social world shifts from family to peers. Duncan, Boisjoly, and Harris (2001) find relatively large correlations among friends on developmental outcomes, in some cases almost as large as correlations among siblings. Peer effects are frequently invoked to understand delinquency and drug use (Akers et al. 1979; Haynie 2001; Sutherland 1947; Warr and Stafford 1991), but l i t t l e recent work addresses the role o f peers in neighborhood effects, particularly with regard to outcomes in other domains. One exception is Anderson (1990,1999), who argues that peer "street" cul- tures in disadvantaged neighborhoods promote teenage pregnancy.

The broader literature on socialization emphasizes that socialization is an active process (Adler and Adler 1998; Corsaro and Eder 1995; Eder andNenga 2003; Simon, Eder, and Evans 1992), in which youths have con- siderable agency to develop "new patterns of t h i n k i n g and a c t i n g " (Corsaro and Eder 1995:433) by mcorporating elements of both the wider adult culture and local peer cultures (Adler and Adler 1998). This "interpretive approach" (Corsaro and Eder 1995) views a youth as a "discoverer o f meaning" (rather than an empty vessel to be filled by adults) who engages in "creative appropriation" through social interaction.

Yet socialization researchers typically study middle-class youth (Adler and Adler 1998; Eder 1995; G i l l i g a n , Lyons, and Hanmer 1990; Simon et al. 1992; Youniss and Smollar 1985), and so they focus either on the family or age- structured settings such as schools or organ-

ized sports (e.g., Fine 1979) where youths tend to develop ties w i t h same-age peers. Socialization of youths in urban neighborhoods, which ethnographers describe as more age and class heterogeneous (Anderson 1990, 1999; Horowitz 1983; Sullivan 1989), is understudied. Perhaps because their subjects tend to be pre- adolescents embedded in age-structured, mid- dle-class contexts, socialization researchers implicitly make a binary distinction between adults and children and assume that such con- texts are the primary settings for formation of friendships and peer groups. Indeed, Eder and Nenga (2003) note the need for research on adolescent socialization outside family and school settings.

The ethnographic and urban poverty literature on cultural transmission processes describes the dynamics of multiple competing sources of socialization among adolescents in disadvan- taged neighborhoods. Classic accounts of crim- inal socialization emphasize the role o f older peers i n introducing adolescents to criminal opportunities and skills (Cloward and Ohlin 1960; Sullivan 1989; Thrasher 1927; Whyte 1943). Anderson's (1990, 1999) ethnographic description o f the competition between con- ventional (decent) and oppositional (street) ori- entations in poor neighborhoods provides the framework for much contemporary research. Decent families compete with peer groups and others with street orientations for adolescents' attention when adolescents make key decisions about school, crime, and sexual behavior. Parents work to separate their children from neighborhood influences (Furstenburg et al. 1999), and schools fail to engage students when teachers interpret their street behavior as resist- ance ( M a t e u - G e l a b e r t and Lune 2 0 0 7 ) . Anderson has been criticized for using his sub- jects' cultural categories as analytical categories (Wacquant 2002), but his ethnographic descrip- tion o f competing cultural codes in poor neigh- borhoods is widely accepted.

Although implicit in many ethnographic descriptions o f life in poor neighborhoods (Anderson 1999; Horowitz 1983; Suttles 1968; Whyte 1943), the causes of cross-age interac- tions and their implications for adolescent socialization are underexplored. Anderson (1990, 1999) laments the decline of the cultur- al authority o f the "old head"—an older man who imparts decent values to neighborhood

4 4 8 A M E R I C A N S O C I O L O G I C A L R E V I E W

boys. Such figures face competition from drug dealers and hustlers presenting alternative mod- els o f success (cf. Young 2007). A l t h o u g h Anderson describes in detail the nature and content of old heads' interactions with adoles- cents, he provides little information about their replacements, the young men held in high regard in the neighborhood because of their resources from the underground economy and their suc- cessful navigation of dangerous streets. How they interact with and transmit cultural models to adolescent boys is not specified, nor are the reasons that adolescent boys take their mes- sages seriously (apart from their social status in the neighborhood). To understand how neigh- borhood cultural transmission processes oper- ate, we need to know how d i f f e r e n t neighborhoods influence adolescent social net- works and how interaction with different peers shapes adolescent cultural frameworks.

Underlying my analysis o f boys' perspec- tives and experiences is a cognitive view o f cul- ture. Rather than examining values or attitudes and explaining individual behavior via involve- ment in a particular subculture in which certain behavior is normative, this perspective views culture as fragmented and composed o f "dis- parate bits o f information and . . . schematic structures that organize that information" (DiMaggio 1997:263). Following Harding (2007), I view poor neighborhoods as cultural- ly heterogeneous, presenting adolescents with a wide array of competing and conflicting cul- tural models. From this perspective, there are multiple cultural models from which to choose (Fuller et al. 1996; Quinn and Holland 1987; Swidler 1986). Key concepts employed to understand socialization are scripts or strategies of action (Swidler 1986), repertoires or toolkits (Hannerz 1969; Swidler 1986), and frames (Benford and Snow 2000; Goffman 1974; Small 2002). A n advantage of this approach is that socialization can occur not just via involvement in a cohesive subculture but also via the intro- duction of alternative ways of conceptualizing problems (frames) and their solutions (scripts, strategies of action). For example, older peers may structure the frames regarding gender dis- trust that adolescent boys bring to romantic relationships (Harding formcoming).

S O C I A L O R G A N I Z A T I O N OF V I O L E N C E

While social isolation theory posits cultural consequences o f neighborhood disadvantage, social organization theory emphasizes com- munity capacity for social control, arguing that neighborhood disadvantage leads to difficul- ties mamtaining order (Park and Burgess 1925; Sampson, Raudenbush, and Earls 1997; Shaw and McKay [1942] 1969). Social organization models have been used primarily to explain neighborhood differences in crime, violence, and delinquency ( f o r an e x c e p t i o n , see Browning, Leventhal, and Brooks-Gunn 2005). Social organization predicts not just a neigh- borhood's level of violence, but also how its residents respond to violence. Sharkey (2006) shows that adolescents in neighborhoods with higher levels of collective efficacy have greater "street efficacy," that is, the perceived ability to avoid violence and victimization in their neigh- borhoods. Neighborhoods w i t h low social organization may have difficulty regulating other adolescent behaviors, but whether social organization theory can explain neighborhood effects on other outcomes remains unclear. This article suggests one avenue through which social organization can have consequences in domains other than violence, such as education and sex- ual behavior. When a community cannot ade- quately control violence, adolescents' social networks, and therefore their socialization, are affected.

Elaborating these neighborhood processes requires understanding the organization of vio- lence in poor neighborhoods. Many studies focus on status contestation in gangs or the interpersonal dynamics of reputation. According to Thrasher (1927), conflict with other gangs is a central element in gang life, and "gang war- fare" erupts over status, economic assets, terri- tory, and the safety o f members. Short and Strodtbeck (1965) argue that gang conflict is also part of status management within a gang, as members use violence within the gang and between rival gangs to establish and maintain leadership roles.

Anderson (1999) describes the social and cultural dynamics of street violence beyond the role o f gangs. Young men "campaign for respect" on the streets according to a "street code" o f informal rules governing masculinity, violence, and public behavior (cf. Mateo-

. . - . 3 * I P

SI

sits cultural isadvantage, asizes com- arguing that to difficul-

rrgess 1925; 1997; Shaw )rganization ' to explain e, violence, p t i o n , see j u n n 2005). rst a neigh- Jso how its rkey (2006) rhoods with rave greater ed ability to their neigh- low social regulating

ether social ighborhood nclear. This vhich social in domains

on and sex- ;annot ade- ints' social ization, are

I processes tion of vio- ny studies ngs or the . According ler gangs is 'gang war- ssets,terri- Short and conflict is

l i n a gang, 3 gang and d maintain

social and beyond the paign for 3 a "street asculinity, f. Mateo-

V I O L E N C E , A D O L E S C E N T B O Y S , A N D O L D E R P E E R S 4 4 9

Table 1. Percent of NELS Respondents from Urban Areas Reporting that They Spend Most of Their Time w i t h People Who Are Older (but under age 26) at First Follow-up (10th grade)

Zip Code Disadvantage Quintile

A l l Adolescents ( N = 9,291)

Male Adolescents (N = 4,514)

Female Adolescents (N = 4,757)

1st (least) 48.5% 42.8% 55.0% 2nd 53.1% 44.6% 60.8% 3rd 57.0% 52.6% 60.6% 4th 58.4% 50.8% 65.5% 5th (most) 64.5% 55.6% 71.5%

All 55.9% 48.5% 62.5%

Source: NELS:88 base year (1988) and first follow-up (1990). Notes: Estimates adjust for the NELS sampling design and weights. In each column, zip code differences are statistically significant atp < .01.

Gelabert and Lune 2007). Where victimization is common, young people view a reputation for toughness—created and maintained by postur- ing and fighting—as a form of protection. As in the gang literature, violence is a means by which status is achieved and maintained. A l t h o u g h this form o f status contestation through violence is often interpersonal, Anderson notes (without development) that it can be group or neighborhood based, as well. Below I build upon these theoretical concepts— status contestation, defended territory, and respect, reputation, and masculinity through violence—to understand the consequences o f violence for the peer networks of adolescent boys in poor neighborhoods.

N E I G H B O R H O O D D I S A D V A N T A G E A N D I N T E R A C T I O N W I T H OLDER PEERS

For older peers to play a role in the differential socialization of adolescents in poor neighbor- hoods, such adolescents must experience greater interaction with them. I investigate whether this is the case using nationally representative data from the National Educational Longitudinal Survey (NELS). In the first follow-up, con- ducted two years after 8th grade respondents were sampled, adolescents were asked about the ages o f the "people with whom you spend most of your time." I use responses to this ques- tion in conjunction with the respondent's age to construct an indicator for whether the respon- dent spends most of his time with individuals who are older than him but under age 26. This age cutoff corresponds most closely to the age

range of older peers described by the adolescent boys in the disadvantaged areas in the field- work presented below.

Table 1 shows the percent of urban adoles- cents who report spending most of their time with individuals who are older but under age 26, stratified by type of zip code, a proxy for neigh- borhood. Zip codes are divided into quintiles of a disadvantage scale constructed from the fol- lowing census variables (measured at the zip code level): family poverty rate, male unem- ployment rate, percent of families headed by a single mother, median household income, per- cent of workers in managerial or professional occupations, and percent of individuals over age 25 who have a college degree. Results are presented for all urban adolescents and for boys and girls separately. Further information about data and methodology is presented in the Online Supplement on the ASR Web site (http://www2. asanet.org/journals/asr/2009/toc069.html).

Three limitations of these data make it d i f f i - cult to detect differences across neighborhoods in the type o f interactions with older peers on which I focus in the qualitative data below (i.e., older, non-kin males in the immediate neigh- borhood). First, zip codes provide very broad definitions o f neighborhoods. Averaging over larger, more diverse geographic areas w i l l atten- uate any neighborhood differences based on neighborhood socioeconomic composition. Second, a key characteristic of disadvantaged neighborhoods in the qualitative analysis is neighborhood violence, but NELS data do not provide information on neighborhood violence. To the degree that some low-violence disad- vantaged neighborhoods are included in the

4 5 0 A M E R I C A N S O C I O L O G I C A L R E V I E W

data, they w i l l attenuate the neighborhood dif- ferences in interaction with older peers. Third, although the qualitative analysis focuses on non-kin older peers, the wording o f the NELS survey question does not allow me to exclude family members from the older peers captured in the NELS data or to focus on same-sex older peers. This may be exacerbated by the "most of your time" wording in the NELS question, which suggests a greater level of interaction than that described in the qualitative interviews. To the degree that kin or opposite sex individ- uals are included in the NELS responses, neigh- borhood differences w i l l again be attenuated. For these reasons, the neighborhood differences observed in the NELS data are smaller than those observed in the qualitative data present- ed below, where the measures can be more pre- cisely defined.2

Despite these limitations, this descriptive analysis provides evidence that, compared with adolescents in middle-class neighborhoods, both male and female adolescents in disadvantaged neighborhoods are significantly more likely to interact regularly with older peers. Although base levels o f older peer interaction are higher for girls than for boys, cross-neighborhood dif- ferences are similar. Girls' greater interaction with older peers is likely due to their dating older boys, a consistent finding in the literature on adolescent romantic relationships (Elo, King, and Furstenberg 1999; Ford, Sohn, and Lepkowski 2001). A logistic regression pre- sented in the Online Supplement controls for a large number of individual and school covari- ates and also finds neighborhood differences in interaction with older individuals.

This analysis simply shows that, although the magnitudes o f neighborhood differences are different, the neighborhood differences in cross-age interaction observed among adoles- cent boys in the fieldwork data also exist in a nationally representative sample of adolescents. It demonstrates an association between neigh- borhood disadvantage and the probability o f cross-age interaction, for which qualitative data are ill-suited.

2 Note that there is also a difference between the time period of the NELS survey (1990) and the field- work (2003 to 2004).

F I E L D W O R K M E T H O D O L O G Y

I conducted unstructured interviews with 60 adolescent boys, ages 13 to 18, living in three predominantly African American areas o f Boston (20 boys per area). For 80 percent, I also interviewed a parent or primary caretaker. I focus on boys because of their greater involve- ment in and exposure to street violence and to allow a gender match between subject and inter- viewer. The Online Supplement provides a detailed description of the fieldwork and analy- sis, including subject and neighborhood char- acteristics, subject recruitment, interview content, data interpretation and cross-check- ing, analysis procedures, and challenges in con- ducting f i e l d w o r k w i t h this p o p u l a t i o n , including issues of race and social distance.

I selected the study areas to allow for explic- it comparisons between similar youth in neigh- borhoods with different poverty rates. Two of the areas ("Roxbury Crossing" and "Franklin") have high rates o f family poverty (35 to 40 per- cent). The third area ("Lower Mills") has a low poverty rate (below 10 percent). Each area con- sists of two contiguous census tracts. Additional neighborhood characteristics from the 2000 Census are provided in the Online Supplement. As I define them, Roxbury Crossing, Franklin, and Lower M i l l s are areas o f the city rather than social neighborhoods. Each area encom- passes multiple locales that more closely approximate neighborhoods recognizable to res- idents, but the neighborhoods within each area share parallel histories and demographic and structural characteristics.

To understand the boys' neighborhoods, the interviews investigated how the subjects con- ceptualize their neighborhoods as geographic and social spaces and how the neighborhood overlaps with their social networks, daily trav- el, and institutions. Although a full discussion is beyond the scope of this article, when the boys and their parents describe their neighborhoods, they are referring to much smaller spaces, often a few blocks in any direction.3 The boundaries

3 Furstenberg and colleagues (1999) note that def- initions o f neighborhood boundaries vary consider- ably across individuals, but I am aware of no prior research that explicitly investigates adolescents' per- ceptions of neighborhood size. However, evidence

Table 2. V i o l City

Homicide

Robbery

Aggravated Ass;

Source: Author's Note: Denomina

of these more! reinforced by Roxbury Cros delineate the anonymity foi ening the ge Hereafter, "ne: graphic and & jects, and "are; study areas.

Space limit description of ence between I low poverty ai zation of violf patterned in u and Raudenbv. 2004). Young borhoods are and violence ( 2000; Center; the adolescei and Franklin their counter] these neighbo present threat

from a differei conceive of the (Cobbina, M i l ! tent with the s Hunter (1974) poor tend to h£ borhoods. Hai social construi more detail.

VIOLENCE, ADOLESCENT BOYS, AND OLDER PEERS 451

vs with 60 ng i n three n areas o f I percent, I y caretaker, .ter involve - ence and to ct and inter- provides a £ and analy- rhood char- , interview •ross-check- nges in con- o p u l a t i o n , distance,

v for explic- itly i n neigh- ;s. Two of the "Franklin")

35 to 40per- s") has a low ich area con- s. Additional m the 2000 Supplement, ng, Franklin, S city rather area encom- ore closely dzable to res- l i n each area Jgraphic and

orhoods, the ubjects con- : geographic eighborhood s, daily trav-

discussion vhen the boys

hborhoods, spaces, often

boundaries

note that def- 'ary consider-

o f no prior olescents'per- ver, evidence

Table 2. Violent Crime Counts and Rates per 1,000 Residents ( i n italics) by Study Area and for City of Boston as a Whole

F r a n k l i n Roxbury Crossing Lower M i l l s City of Boston

2003 2004 2003 2004 2003 2004 2003 2004

Homicide 4 8 1 ' 0 0 3 39 61 .34 .67 .16 .31 .07 .10

Robbery 16 68 48 46 26 23 2759 2433 6.39 5.71 7.78 7.46 2.72 2.40 4.68 4.13

Aggravated Assault 171 157 83 90 58 44 4113 4151 14.37 13.19 13.46 14.60 6.06 4.60 6.98 7.05

Source: Author's calculations from incident data provided by Boston Police Department. Note: Denominator for rates is 2000 Census total population.

of these more limited spaces are to some degree reinforced by the violence described below. Roxbury Crossing, Franklin, and Lower Mills delineate the three study areas and provide anonymity for the research subjects by broad- ening the geographic scope o f reference. Hereafter, "neighborhood" w i l l refer to the geo- graphic and social spaces defined by the sub- jects, and "area" w i l l distinguish the three larger study areas.

Space limitations preclude a more detailed description of the three areas, but a key differ- ence between the two high poverty areas and the low poverty area is the level and social organi- zation of violence. Violence is highly spatially patterned in urban areas (Morenoff, Sampson, and Raudenbush 2001; Sampson and Morenoff 2004). Young people from high poverty neigh- borhoods are exposed to high rates o f crime and violence (American Academy of Pediatrics 2000; Centers for Disease Control 1997), and the adolescent boys from Roxbury Crossing and Franklin are no exception. In contrast to their counterparts in Lower M i l l s , the boys i n these neighborhoods must contend with an ever present threat o f conflict and victimization.

from a different city also suggests that adolescents conceive o f their neighborhoods as particularly small (Cobbina, Miller, and Branson 2008). Also consis- tent with the small neighborhoods identified here, Hunter (1974) finds that young individuals and the poor tend to have smaller definitions of their neigh- borhoods. Harding (forthcoming) investigates the social construction o f neighborhood boundaries in more detail.

Table 2 shows violent crime rates and counts for the three study areas and for Boston as a whole during the fieldwork period. Because they are rare events, homicides provide indica- tions of the most severe violence. By this meas- ure, Franklin is the most violent of the three study areas. Rates of robbery and aggravated assault, much more frequent events, illustrate the key distinction between Lower M i l l s , the low poverty area, and the two high poverty areas, Roxbury Crossing and Franklin. By both meas- ures, Roxbury Crossing and Franklin have much higher rates of violence than do Lower Mills or Boston as a whole. This street violence is rarely random or indiscriminate. Rather, it is organized by both interpersonal and neighborhood-based "beefs," ongoing conflicts that wax and wane and are largely based on individual and neigh- borhood status contestation. Neighborhood rivalries structure much of the serious violence that occurs in Boston's disadvantaged neigh- borhoods, including Roxbury Crossing and Franklin.

The ability to make comparisons across neighborhoods is a key aspect o f this study's design. Few qualitative studies allow systemat- ic comparison o f experiences in poor and non- poor neighborhoods. Asking similar questions and discussing the same topics with individu- als i n different neighborhoods reveals key dif- ferences in the daily lives of adolescent boys across neighborhoods. It was only through these explicit comparisons that I discovered neigh- borhood differences in experiences o f violence, threats of victimization, and the role of older males i n social networks. Across areas, it is important to compare adolescents from similar

'•V

452 AMERICAN SOCIOLOGICAL REVIEW

family backgrounds, so I selected subjects to achieve economic diversity in each area.

A G E STRUCTURE OF PEER N E T W O R K S

Compared w i t h their counterparts i n Lower M i l l s , the boys i n Franklin and Roxbury Crossing interact more often with older ado- lescents and young adults, particularly those from their own neighborhoods. As suggested by the NELS data, adolescent boys' peer networks in these two disadvantaged areas include indi- viduals who are considerably older. Of the boys interviewed in Franklin and Roxbury Crossing, 75 percent reported older men from outside their families as part of their peer networks. These older men were at least two years older and sometimes in their mid-20s. By contrast, only 15 percent of Lower Mills subjects report- ed interacting with older adolescents and young adults outside their families. In this section, I describe how the social organization o f vio- lence structures the age composition o f ado- lescent boys' peer networks i n R o x b u r y Crossing and Franklin. In highlighting the role o f .same-sex older peers, I am not arguing that these actors are the only source of socialization for the Franklin and Roxbury Crossing sub- jects. Parents, siblings, same-age peers, girl- friends, and extended families play important roles as well. I focus on older peers because their role has not been previously explored, because comparisons across neighborhoods revealed this stark difference, and because of their poten- tial role in non-family cross-cohort socialization.

Two mutually reinforcing processes explain why the adolescent boys o f F r a n k l i n and Roxbury Crossing tend to associate more with older peers. First, neighborhood violence increases the salience o f neighborhood identi- ty, leading boys to choose friends from those available locally. Boston boys view their neigh- borhoods as small geographic areas. Venturing outside the neighborhood or interacting with boys from other neighborhoods increases the risk o f victimization, so fewer same-age friends are available and use o f geographic space is constrained. Second, as the younger adolescent boys o f Roxbury Crossing and Franklin strug- gle to cope with the ever-present threat o f vio- lence, relationships with older peers are one strategy for securing at least a measure o f pro-

tection. In times o f danger, older peers can be called on to intervene, and relationships with older peers provide status and reputational advantages that can help prevent conflict. By contrast, adolescent boys in Lower Mills face a much lower threat of victimization. Their social lives are not structured by strong neighborhood identities that restrict social networks or the use o f geographic space, so strategies for reducing the threat of victimization are less necessary.

NEIGHBORHOOD VIOLENCE AND SOCIAL IDENTITY

Among the adolescent boys o f Franklin and Roxbury Crossing, neighborhood violence is simultaneously structured by neighborhood identities and a primary way in which these identities are experienced and reinforced in daily life. Nearly all o f the subjects in these two poor, violent areas use neighborhoods as categories to distinguish insiders from outsiders. Neighborhoods strongly influence, but do not rigidly determine, which male peers boys social- ize with and would consider as potential friends. In Franklin and Roxbury Crossing, 32 o f 40 boys (80 percent) reported that more than half o f their friends live in their immediate neigh- borhoods, and 22 of the boys (55 percent) had no friends from outside their neighborhoods. Only three of 40 boys reported no neighborhood friends. In Lower Mills, 13 of 20 boys (65 per- cent) reported that more than half of their friends live in their neighborhoods, and seven of the boys (35 percent) had no friends outside the neighborhood. Four of 20 Lower M i l l s boys reported no neighborhood friends.

M a r c u s , a 16-year-old f r o m R o x b u r y Crossing, helps illustrate the power of neigh- borhood distinctions. A boy from Marcus's development owed him four dollars. Marcus wanted the money but did not want to fight his neighbor for it. Instead, Marcus told him that i f he did not repay, he would rob his friend from another neighborhood:

A kid owed me $4 He lives in this development, so I didn't want to do nothing to him, because I knew it would cause problems. I said, " I see you coming around here with a kid that you hang out with, and I really don't care for the kid, 'cause the kid's not from around here. It's either I get my money, or we're going to have problems," because that $4 means something to me, 'cause my fami- ly does not have that much. I told him, " I ' l l rob your

VIOLENCE, ADOLESCENT BOYS, AND OLDER PEERS 4 5 3

peers can be mships with reputational conflict. By

r Mills face a .. Their social leighborhood rks or the use for reducing necessary.

SOCIAL

Franklin and d violence is leighborhood

which these reinforced i n jects in these hborhoods as

m outsiders, but do not

boys social- Qtial friends.

, 32 o f 40 than half

sdiate neigh- percent) had "hborhoods.

ghborhood >oys (65 per- ? their friends seven of the outside the M i l l s boys

R o x b u r y o f neigh- Marcus's . Marcus fight his

urn that i f end from

'elopment, because I I see you hang out cause the I get my 'because my fami- rob your

friend the next time I see him around here." . . . I had my $4 in my hand the next day.... We just shook hands, and we left it at that.

This story reveals the salience o f neighbor- hood distinctions for Marcus, but it also indi- cates that this distinction is not universally important, as the other boy did have a friend from another neighborhood. From Marcus's perspective, conflicts among neighborhood insiders are different from those between youth from different neighborhoods. Conflicts among neighbors are resolved quickly, although some- times with a physical fight. Third parties from the neighborhood take on a mediating role to resolve a dispute before it escalates into an ongoing beef. There is often social pressure to resolve a dispute and "leave it at that." When conflicts between youths from different neigh- borhoods emerge, however, each youth becomes a representative o f his neighborhood, with its reputation at stake. Others may become involved, either to seek retribution and redemp- tion for the neighborhood or to protect its rep- utation.4

Consider an incident recounted by Chris, a 14-year-old Franklin resident. Soon after arriv- ing at a nearby roller rink, Chris and "his boys" encountered a group o f youths from a rival street who insulted a girl from his neighborhood:

We almost got into a fight. Some Lucerne [Street] kids was there.... They were just making mad noise. Cause in the Vous [the roller rink] they're like "Lucerne, Lucerne." . . . There was all kinds o f kids there though. The Point was there, and the Head was there. D Block was there.... They had a dance contest and the girl that was from [our neighborhood], she was dancing. And they was talking mad trash, so we almost got in a fight with them. They was scared. They l e f t . . . . We was there like 20 deep.

Neighborhood is a central organizing cate- gory in Chris's account. It structures not only what happened, but the way he tells the story. In mentioning the Point, the Head, and D Block, Chris describes other youths using local slang

4 The role o f third parties in within- versus cross- neighborhood violence is consistent with Black's (1993) argument that vengeance increases with social distance (Phillips 2003) and the finding that medi- ating third parties within groups can reduce violence (Phillips and Cooney 2005).

for their neighborhoods. Chris and the boys from his neighborhood are not being chival- rous i n defending the g i r l . Rather, they are defending their own neighborhood from insult with a masculine display of physical force. A n outside observer might expect an older brother or boyfriend to challenge the insult, but here the conflict quickly became a contest between neighborhoods rather than individuals.

W i t h these dynamics, neighborhoods i n Roxbury Crossing and Franklin become what Suttles (1972) terms "defended communities." Neutral spaces such as schools, public trans- portation, and commercial areas are also sites o f contestation and conflict. Confrontation between youths from different neighborhoods is always a possibility, and youths often ask each other what neighborhood they are from as a challenge or physical threat. I n Roxbury Crossing and Franklin, violence, neighborhood identity, and community membership are close- ly linked. Tyree, who is 17 and from Roxbury Crossing, described how he experiences the obligations o f community membership with regard to violence:

I f you're not w i l l i n g to help in the neighborhood, then you really can't be here. Like, i f you wasn't in that circle, you was outside that circle There's people in the neighborhood that live on the same street and could see you getting jumped that would- n't care. "Oh it's not my problem." But i f you was really tight, grew up together, been through ups and downs and really been cool, know family members, then there's always a chance of help. I always got to help the people in my neighborhood because you never know when it's going to come back to you. I could be outside the neighborhood getting ready to get jumped, and he could walk by and help.

People who do not support their neighbors in fights or do not defend the neighborhood can- not rely on neighbors for other forms of assis- tance. Although not all boys see things this way, for adolescents like Tyree, participating in this system of obligations defines membership in the community, including access to mutual protec- tion.

By contrast, the boys of Lower Mills do not link identity and community membership to mutual obligations of protection. Lower M i l l s boys tend to look to sources of membership other than the neighborhood to situate their identities. Small friendship groups, interests such as sports or music, and involvement in religious or ethnic groups are more important.

4 5 4 AMERICAN SOCIOLOGICAL REVIEW

For example, Isaac, age 13, is the quarterback of the Pop Warner football team. He frequent- ly plays football at either the park near his house or the after-school program he attends. He has one or two friends from the neighborhood but most are from his team or program. He describes himself as someone who does not "have any enemies," by which he means that, unlike oth- ers at his public middle school, neither he nor his neighborhood has beef with anyone.

Jason, who is 18 and from Lower Mills, also confirms the absence o f neighborhood-based identities and rivalries in Lower M i l l s . He dis- missed the notion of gangs in his neighborhood as i f gangs were a passing phase among younger adolescents:

Gangs are kind of like 8th grade. That's kind of dry now. Nobody is in a gang any more. You do what you gotta do. I know people that is like freshman year in school, don't see them wearing their ban- dannas any m o r e . . . . Like you won't hear anyone like, "Oh yeah, a whole bunch o f Crips jumped some old lady" or something. It wouldn't be like a gang, it would just be a group of kids. It's not real- ly a gang any more cause when it comes down to it, people are mostly for theirself.

For Jason, group-based conflicts and the social symbols that go with them (colored ban- dannas) are a middle-school fad, lifted from popular culture. Some remnants remain, but Jason does not see such conflicts affecting his daily life; violence does not touch Lower Mills youths like Isaac and Jason very often. When violence does occur, it is both less serious and more likely to stem from interpersonal disputes than from ongoing neighborhood conflicts. I f a boy from Lower M i l l s ventures outside his neighborhood, he must avoid encroaching on others' turf, but he need not fear encountering someone with whom his neighborhood has beef. Lower Mills neighborhoods are outside the sys- tem of neighborhood conflicts.

Parents also take seriously the connections between safety, neighborhood identity, and con- flict. For example, two parents cited neighbor- hood identities as affecting their decisions about school choice. When Terrell, from Franklin and now age 16, and his mother tried to decide which high schools he preferred, she insisted that he not include two schools on his preference list because many youths from a rival housing development attend them, which would put him at risk of assault at school. When Manuel, now

age 15, and his mother made the same decision, she favored a particular high school because other youths from their Franklin housing devel- opment attend that school. He would have neigh- borhood friends to keep him safe at school and while traveling there and back.

By contrast, instead of violence, Lower Mills parents worry about other challenges that ado- lescent boys face, like staying focused on school and avoiding early fatherhood. Darnell's moth- er, for example, focuses on her fear that her 13- year-old son w i l l get "sidetracked" from school:

Violence? This neighborhood, not too much. Most of the kids, they basically play at the park and my son is not the fighting type. That is not one o f my concerns, at least not right now.... I don't want him to get sidetracked. Girls, no. Girls and falling into the wrong crowd. I just think i f he sees people doing things that they think is fun but is stupid, that he might want to try it too, just cause everybody else is having f u n . . . . I mean because he is very independent. He spends a lot o f time by himself just by choice. Like sometimes I have to tell him, go outside and play, but he is in his room writing and reading I think he's getting into girls. Once he really gets into girls, I try and steer him away from that, sex and getting girls pregnant, diseases and just all the little things that can go on between teenagers.

While Franklin and Roxbury Crossing parents worry constantly about their sons becoming enveloped in the neighborhoods' dangers, Darnell's mother makes it a point to encourage him to inter- act with his neighborhood friends and can focus her parenting energy on other concerns.

The link between neighborhood identity and violence affects not only those actively involved in violence but also the boys who remain on the sidelines. Only a small proportion of Franklin and Roxbury Crossing youths defend their ter- ritory from intrusion by others, confront youths from other neighborhoods in neutral territories, or carry out the retribution and revenge that keep beefs going. Some boys even opt out of neighborhood social life altogether (Furstenburg et al. 1999). Nonetheless, any boy who ven- tures outside his own neighborhood to go to school, a store, or downtown risks confrontation with youths from other neighborhoods. Most youths travel in groups to avoid being harassed when leaving their neighborhoods. Even those Franklin and Roxbury Crossing youths who never start fights with youths from other neigh- borhoods are enveloped in the system of place-

; decision, )1 because ling devel- lave neigh- school and

ower Mills s that ado- 1 on school ell's moth- hat her 13- om school:

much. Most )ark and my )t one of my n't want him I falling into sees people s stupid, that 5 everybody e he is very : by himself ; to tell him, 3 0 m writing 3 girls. Once er h i m away ant, diseases i on between

;ing parents becoming

rs, Darnell's rim to inter- d can focus ns. dentity and :ly involved main on the of Franklin id their ter- r ont youths : territories, :venge that opt out o f

^urstenburg ' who ven- )d to go to nfrontation Dods. Most lg harassed Even those ouths who ither neigh- m ofplace-

VIOLENGE, ADOLESCENT BOYS, AND OLDER PEERS 455

based antagonisms. Terrell was threatened with a gun when he visited a friend who lived near a rival housing development. The two develop- ments had beef at the time, but Terrell was never involved until then:

Just from us living around here, sometimes it's a safety issue... . Because these people [from our development] get a lot of people riled up against them. So they want revenge in any way. And they don't care i f you hang with them or you don't hang with them, as long as you live around here, you're a target to certain people.

I was actually with one of my friends, and we were going to [a grocery store] in Jamaica Plain. We were walking up and some people asked me where I was from, so I told them. And then, they pulled out a hand gun on u s . . . . It had to be like three o'clock in the afternoon.

And it's more than just trying to ignore it, you gotta watch your back too. You can't just say, "Well yeah, I ' m from around here, but I don't mess with those guys [who are involved in neighborhood beefs]."They're gonna say, "So what!"

Contrast Terrell's experience w i t h that o f Delbert, age 14 and from Lower Mills. Like his Franklin and Roxbury Crossing counter- parts, Delbert tries to travel with a group when he leaves his neighborhood, even to take the train to go downtown. Because his Lower Mills neighborhood is outside the system o f neigh- borhood beefs, however, his chances of victim- ization are far lower than Terrell's:

I was on the T [the Boston subway] one day. I was coming from downtown, the m o v i e s . . . . And I was sitting there and these three kids who were like 17 came over to me and was like, "Where are you going, where are you from, do you know the big gangs that are around there," and questions like that and I answered no to all o f them.

Int: And then they left you alone? Yeah, cause I said no.

The youths who confronted Delbert were looking for people from certain neighborhoods and approached him because they suspected he was from one o f them. As a resident o f Lower M i l l s , however, he has no connection to the neighborhood conflicts that the youths on the train were trying to draw out, and so he was left alone. Neighborhood identity does not structure Delbert's safety as it does for Terrell.

NEIGHBORHOOD IDENTITIES AND CROSS-

AGE INTERACTIONS

The small size o f the neighborhood territories amplifies the consequences o f neighborhood identities and defended territory for an adoles- cent boy's freedom of movement. Boston's rel- atively small public housing developments, usually covering only a few blocks and con- sisting of several hundred families, are natural organizing units. Large private or co-op hous- ing developments are also natural geographic areas. Many of the geographic areas, however, are single streets, one to three blocks long, and sometimes parallel streets have longstanding beefs.

Terrell's experience illustrates how simply being associated with a particular neighbor- hood enmeshed in ongoing beefs can limit one's freedom o f movement. Many adolescents in Franklin and Roxbury Crossing adopt the sur- vival strategy of remaining in their home neigh- borhoods as much as possible, geographically constricted as it may be, and avoiding interac: tion with males from other neighborhoods.5 Simon, a 16-year-old Franklin resident who has experienced many altercations with adolescents from other neighborhoods, rarely leaves his neighborhood anymore. He explains how leav- ing can easily lead to a violent confrontation i f he is not with a group o f friends:

I don't feel safe at all. Well, probably a little bit, like in my neighborhood, it's all right. But other places? Noooo, I don't really go out of my area that much because I know people, just from looks, get murdered.... Because you're like an alien, you're not known over there. So the first thing you do when you walk through there, all eyes is on y o u — "Where you from?" . . . I f they got problems with that area They just see you over there, and knowing you're not from over there, and they just set it off with you right there It happens to me a lot of times The best thing to do, i f you're not with people you know, you have to be at the [local community center], because that's the most safest thing. That's about it i f you're not with a lump sum [large group] o f people.

5 Cobbina and colleagues (2008) also note that ado- lescents in high violence areas view their immediate neighborhood as safe and stay close to home as a safe- ty strategy.

4 5 6 A M E R I C A N S O C I O L O G I C A L R E V I E W

This strategy leaves only people who live in one's own neighborhood as potential friends. The social space in question is quite small, so same-age peers are often not abundant, leaving older peers to f i l l the gap. Marcus, age 16, described how the lack of same-age peers in his development led him to socialize with the "older guys" who are a fixture of the streets. These older guys dispense advice to Marcus and his friends:

[ M y two friends and I ] are the three younger indi- viduals that live around here, so we're forced to be around nothing but older guys. . . . We're put around older guys that done been through it, that tell you what to do, and what not to do, and how to do it, and when to do it, and when it's appropriate to confront somebody, or that you've got a prob- lem with.

We just hang out, talk about things that done happened—laugh, joke with each other. One per- son might be fixing on their car; one person might be fixing on their bike, listening to music, and we all just go gather around there, and just t a l k . . . . Maybe the older guys might be sitting down, and drinking beer and playing cards, and everybody just gather around.

The scene Marcus described is frequently visible to any observer who spends time in the neighborhoods o f Franklin and Roxbury Crossing. Hanging out in a mixed age group (with older individuals at the center), usually on the basketball court, a stoop, or the corner, older adolescents and young adults recount their expe- riences and dispense general advice.

Previous ethnographic work describes the extensive use o f public space in poor neigh- borhoods, owing to overcrowded apartments and lack o f air conditioning in the summer, which leads to "street interaction and informal meetings" (Horowitz 1983:39). Similarly, for adolescent males i n Franklin and Roxbury Crossing, parks and corners are a venue for cross-age interaction. Joseph, age 15 and from Franklin, spends most of his non-school time at the neighborhood basketball court and nearby athletic fields, where he encounters individuals o f every age, some older and some younger:

I know little kids and older, like adults, like in their 20s or so. Teenagers my age. I know most everyone on that court. And we all just play with them. And now we play with our own age and stuff like that. So it carries on through the ages, like a legacy.

For Joseph, the neighborhood's public spaces are where the "legacy" is passed down to him from the older generations, and where he w i l l pass it on to the younger boys he encounters as well. Limited from venturing outside their local area by vio- lence, boys like Joseph come into frequent contact with older males in these settings, and these males become part of their social networks, not neces- sarily as friends but as acquaintances.

The streets and parks of Lower Mills neigh- borhoods present a stark contrast. Compared with Franklin and Roxbury Crossing, the reg- ular observer of Lower Mills will witness far less activity in the neighborhood's public spaces. Greater family economic resources mean more youths participate in formal programs after school and on the weekends and have the pock- et money to venture outside the neighborhood to the movies or the arcade. For those without such resources, there are simply fewer others of any age hanging around the neighborhood. When the youth of Lower Mills do associate with older peers, their interactions are more often family-based and take on a different char- acter. When age inappropriate discussions or activities arise (e.g., concerning romantic rela- tionships, sex, or drug or alcohol use) younger adolescents are pushed away. Delbert described his experiences with older neighborhood peers:

Int: You said the older youths would be your broth- er and his friends?

Yeah. His friends about like a range o f 17, 18, and 19. Some ofthemare20 They don't let us hang out with them. And then i f we try hangin' out with them, they'll say get out of here.

Int: Why? Because they don't want us to do bad things.

They want us to live our own lives and not copy what they do and to be ourselves.

Int: What kind of bad things might you be doing? Smoking [marijuana] probably. Int: And why would you want to hang out with

them ? Cause they're older and they're more experi-

enced and some of them . . . [pause], but most of them look like they're going in the right direction.

Just as importantly, the working- and lower- middle-class neighborhoods of Lower M i l l s have fewer idle young men and older adoles- cents of working age with whom younger ado- lescents might pass the time.6 By contrast, Tyree,

6 See the Online Supplement.

lie spaces are him from the pass it on to

veil. Limited area by vio- l e n t contact i these males s, not neces- s. Mills neigh- :. Compared ing, the reg- itness far less iblic spaces, s mean more >grams after ive the pock- eighborhood rose without wer others o f ighborhood. do associate ns are more ifferent char- scussions or imantic rela- use) younger >ert described )rhood peers:

beyourbroth-

nge o f 17, 18, ley don't let us try hangin' out

i o bad things, i and not copy

'you be doing?

hang out with

more experi- : ] , but most o f ight direction.

I - and lower- Lower M i l l s ilder adoles- /ounger ado- mtrast, Tyree,

VIOLENCE, ADOLESCENT BOYS, AND OLDER PEERS 457

now age 17, described the characteristics and behaviors of the older peers in the Roxbury Crossing neighborhood where he grew up:

I'd say they were like 16 to like 22. They was basi- cally not going to school, selling drugs, shooting guns, and really basically doing all the negatives. Like they would rather sell drugs than go job hunt- ing. . . . They rather stand outside on the comer all day than make sure their little brothers and sisters is getting off to the school bus.

It seems like the older you get the worser it gets. Like i f you're living in the neighborhood, i f you started off young, the older you get, the hard- er it is to leave the neighborhood because you feel that's where you're from and that's what you got to, you got to hold the neighborhood down.

As a younger teen, Tyree did not see older peers who were working or in school because they were not hanging around in the streets. As a result, the older males who were actually pres- ent seemed more and more disadvantaged as they got older. Tyree began to see himself becoming more like them as he spent time with them and became involved in neighborhood rivalries, what he calls "holding down" the neighborhood.

It is clear why younger adolescent boys grav- itate toward the available older adolescents and young adults—for both status and, as w i l l be dis- cussed below, protection—but why do older adolescent males and young adults in Franklin and Roxbury Crossing attend to younger teens? First, they see it as their duty to look out for the younger adolescents and children in the neigh- borhoods with which they so strongly identify, similar to the "fictive k i n " relationships Stack (1974) describes. They learned this behavior from adults and from the youths who came before. They also see this guardian and friend- ship behavior as a way to keep the neighborhood safer, to keep the peace and prevent the drama that conflicts can create. David, a 17-year-old from Franklin, described how he watches over younger adolescents as part o f his role as a community member. Others who came before looked out for fhm, and the youths he looks out for w i l l grow up to do the same:

I f I look after somebody, then, they can probably grow up and look after somebody else—that's basically what makes it a community—everybody looking after somebody. So, i f something happens in your neighborhood you can like stop it before it gets worser. Like somebody selling drugs, shoot- ing. Before those folks moved in, there was none

of that stuff going down, we could pinpoint it right now—like take that out of here. And, basically, just watch out for your kids. .

Note that David refers to this as watching out for "your kids," indicating his identification with the neighborhood as a locus o f cross-age ties. Yet caring for younger children and ado- lescents extends beyond simply keeping the peace. Older adolescents recount with pride how they help the younger ones by giving them a dollar when the ice cream truck arrives or when they want a snack from the corner store. As Stack (1974) and Newman (1999) describe, those who have money or other resources are obliged to share it with others in the communi- ty, and these reciprocity norms extend to ado- lescents as well. For Marcus, relationships with younger adolescents have a particularly kin- like motivation: the lack of fathers among neigh- borhood youth.

Some kids don't know who their father i s . . . . I never had a big brother, so I never had an older brother to look up to, to throw a football around with, because me and my father never really did that So, when a younger guy comes and asks me to play ball, I ' l l be happy to play with him, because I want him to feel like he has an older per- son to hang out with.

Int: So what are some of the things that you try to impart on the younger kids?

Just like the older guys tell me. I tell the little dudes that are maybe 13, 12, "Don't hang out in one spot for too long, because that's going to make it look bad on you, and then it's going to make it look bad on us, because now they [the police] think that we have you out here doin' stuff, so now they're goin' to think that we're trying to start up something."

As Marcus's account suggests, an older ado- lescent's interactions with younger adolescents are about more than a simple altruistic notion o f community and collective responsibility. By taking responsibility for their juniors, older peers gain the status that comes with such roles. Whether providing a dollar for the ice cream truck, offering instruction on how to shoot a bas- ketball, giving a lesson on avoiding police harassment, or sharing knowledge about roman- tic relationships, Marcus gains the respect of both the neighborhood kids and others his age who play the same role. Although the focus of this article is on the potential negative—and often unintended—consequences o f cross- cohort socialization, these accounts illustrate

458 AMERICAN SOCIOLOGICAL REVIEW

that cross-age interactions are often altruistically motivated and serve constructive functions as well.

OLDER PEERS AS A PROTECTION STRATEGY

Despite the power o f neighborhood identities and conflicts to limit the mobility and social net- works o f adolescent boys in Roxbury Crossing and Franklin, the boys must leave the neigh- borhood sometimes, at least for school. In addi- tion, although cross-neighborhood beefs are a key part of youth violence in Boston, conflict among youths from the same neighborhood also occurs. For both these reasons, adolescent boys in Franklin and Roxbury Crossing must develop protection strategies. As discussed above and in prior research, these strategies include traveling in groups, staying close to home, developing a reputation as a tough fight- er, and avoiding particular people, locations, and activities that might increase one's risk o f victimization (Anderson 1999; Cobbina et al. 2008; Jones 2004). Many youths employ mul- tiple strategies, while others withdraw from neighborhood social life entirely (Furstenburg et al. 1999), as did several o f the Roxbury Crossing and Franklin subjects.

Another strategy involves relationships with older peers who can provide security through both direct intervention and reputational sta- tus. A n example of older peers providing a measure of protection comes from Chris, age 14, whose Franklin housing development has an ongoing beef with a nearby rival development. As Chris explained, "the drama" began again at a party:

Me and my boys bagged these Franklin Field girls [got their phone numbers], and the guys from Franklin Field got mad. So they brought it [the fight] to us. We beat them up a little bit. Then the older mens came, then we got our older mens, so that it looked like a go. But then the girls called us like, "We don't want drama." So I was like, forget it.

As Chris made his way home, however, he was jumped by Franklin Field youth riding in a car driven by one of the "older mens." They were looking to win back some pride for their neighborhood. As his mother confirmed, Chris was beaten fairly badly but suffered no serious injuries. As the cross-neighborhood conflict escalated, Chris and his friends secured the help

of some older adolescents and young adults (the "older mens") from his development. This show of force seemed to halt the conflict:

Then they [the youth from Franklin Field] started riding through here [Franklin H i l l ] . So me and my little niggas told our older niggas, and then they got involved. Once they got involved, they [Franklin Field] didn't want it. Our OG's [original gangstas], they didn't even get involved. There was just me and the older niggas. It was just a cou- ple of niggas from up here [Franklin H i l l ] , and they [Franklin Field] couldn't handle i t . . . . Because i f the whole [Franklin H i l l ] team came together, it would have been a problem.

Three sets o f actors o f different ages and experiences populate Chris's account. "Little mens" (or "little niggas") are neighborhood teenagers close to Chris's age (about 13 to 16 years old). "Older mens" are in then late teens or 20s and, as veterans o f previous beefs, have honed their reputations based on past deeds. Currently, they may be hustling or dealing drugs or working, but they are usually a common presence on the streets. OG's, or original gang- sters, are in their 30s and 40s. survivors of the worst days o f violence in Boston. They grew up in the neighborhood and were leaders in their day. Their sometimes legendary reputations cement their place at the top o f the social hier- archy, but their daily connections to the neigh- borhood are often weak, as many have moved away from street life. Chris's account illustrates the complexity of social interactions with older peers. The older peers' role in cross-neighbor- hood conflicts can vary, in this case they helped to contain it.

James, who is 14 years old and from Roxbury Crossing, provides another example of how relationships with older youth provide protec- tion in a violent neighborhood:

I know mostly everyone cause I've grown up with everyone. So kids who are 11 and I was 7 and 8, I was probably outside playing basketball with, or they seen me. So now they're probably about 16, 17.... The kids who was doing bad things, I know most of them. So I wouldn't feel bad walking down the street cause they probably wouldn't mess with m e . . . . I know most of them with a good relationship; I could talk to them and hold a con- versation, but that's about i t . . . . I f somebody was like trying to jump me or do something stupid, they'd probably come in and stop it.

In addition to direct intervention, an associ- ation with older friends can provide status and

VIOLENCE, ADOLESCENT BOYS, AND OLDER PEERS 459

ling adults >ment. This mflict:

ield] started So me and

md then they Dived, they G's [original lived. There isjustacou- ill], and they . Because i f : together, it

t ages and int. "Little ghborhood i t 13 to 16 r late teens )eefs, have >ast deeds, i l i n g drugs i common ;inal gang- vors of the ly grew up :rs i n their 5putations ocial hier- the neigh- ive moved illustrates w i t h older •neighbor- ley helped

n Roxbury le o f how de protec-

•wn up with' •as 7 and 8, iall with, or y about 16, Qgs, I know id walking uldn'tmess ith a good hold a con- lebody was ing stupid,

an associ- status and

respect, which translate into protection. When others know that a boy has older friends, that alone is often enough to make them think twice about "messing w i t h " him (see Jones 2004:55 for a similar example among adolescent girls). Miguel, age 16 and from Roxbury Crossing, described the benefits of his relationships with older youths:

Most of my friends are a lot older than me. Probably like, maybe 20.

Int: How did you get to know them? Hanging out, like in the summer time we go out,

we just sit down on the steps and chill, we talk and stuff like that. People playing baseball, they just tend to come over and they start playing and you just get to know them.

Int: So why would someone who s younger want to be friends with people who are older?

A lot more people will respect the younger per- son. They wouldn't mess with him because he has a lot of older friends.

Miguel's interactions w i t h older youths, which started when he moved to the housing development at age 13 and began hanging out with them, earn him respect from others. In Miguel's view, the respect that these relation- ships provide can protect him. When retaliation from older youths may be forthcoming, ado- lescents may be less likely to start a fight.

I have emphasized the importance of neigh- borhood violence in understanding the greater cross-age social interaction in Roxbury Crossing and Franklin compared with Lower Mills. Family connections, however, could be an alternative explanation. Lareau (2003) argues that among lower-class families there is considerable inter- action with extended family members. Although the data from this study are consistent with this description of such families, the interactions with older peers described by adolescents in Roxbury Crossing and Franklin were not typically facili- tated by family connections. Nor is there a dif- ference between adolescents with large extended families and those without such families.

I M P L I C A T I O N S FOR S O C I A L I Z A T I O N

I have argued that both the level and the organ- ization o f violence in Franklin and Roxbury Crossing increase adolescent boys' interactions and relationships with older adolescents and young adults. I now discuss the potential impli- cations of these interactions for socialization. The strategies that the boys of Roxbury Crossing

and Franklin use to deal with violence and vic- timization can have unintended consequences. Although relationships with older peers pro- vide some protection, they also expose these boys to alternative or ghetto specific (Wilson 1996) cultural models that influence their behav- ior and decision making in other domains, such as romantic relationships and schooling. By contrast, Lower Mills boys tend to have more age homogenous friendship groups. Their peer networks are more similar to those of middle- class children in the bureaucratized and age- graded social settings that Lareau (2003) describes. As a result, older peers play a small- er role in the socialization of Lower Mills boys.

The potential consequences of interactions with older peers depend on their characteristics. Not every older adolescent or young man in the neighborhood sells drugs, impregnates mul- tiple partners, or drops out of school. But indi- viduals who might serve as more positive role models are not as visible because they spend more time outside the neighborhood, either working or in school. The older peers who are available and visible in Roxbury Crossing and Franklin, and who can provide some measure of protection, are seldom positive role models, although not for lack o f good intentions. Marcus, who tries to be an older brother figure to the boys in his neighborhood, is an occa- sional drug dealer, frequently skips school, and at age 16 already has two young women who claim they are pregnant by him.

Eduardo, now age 17, recalled how when he was 12 years old, older friends in the neigh- borhood introduced him to drinking, marijua- na, and stealing cars. A t about the same time, he became considerably less interested i n school:

I turned 13 in fall of '99. That summer of 2000, that's when everything went downhill. By the time 8th grade came around, I went to school for maybe about a week.

Int: Why do you think you weren't interested in school?

The peers around me. Smoking [marijuana], drinking. We were always around girls. Those type of things. Stealing cars. Selling car parts The so-called friends around me, they knew how to steal cars. I just got into it. At first, I just started breaking car windows, then I learned how to steal a car.

I considered them my age. Now, that I look back at it, though, they weren't my age, but when

460 AMERICAN SOCIOLOGICAL REVIEW

I thought about it [at the time], they were my age.... 'Cause my whole life I grew up around people a lot older than me, and that affected me a lot The other guys that were 14 or 15 were still acting younger. At 12,1 was into girls. I was into partying.... I was around older guys.... I just ended up acting mature. The other 12-year-olds, they were thinking about, mommy, buy me some sneakers. At 12-years-old, I was thinking how to get 'em myself.

Eduardo imitated the behavior of older friends with whom he was spending time, but older friends and acquaintances can also have more subtle influences on younger adolescents through then cultural power to frame and contextualize daily life i n the neighborhood and beyond. Through both then words and their deeds, they expose younger adolescents to and reinforce cul- tural models that often differ from those privi- leged i n middle-class culture. The older adolescents and young adults who are respected in Roxbury Crossing and Franklin because of their mastery of the streets regularly dispense advice about girls, school, and staying out of trouble. The guidance comes in the form of gen- eral statements that can be interpreted and applied to future situations or from recounting experi- ences and interpretations of daily life.

Such general statements have little impact when they come from teachers or parents. But older adolescents and young men, especially those with status earned on the streets, command the attention o f their younger counterparts because they have been through experiences that the younger boys w i l l soon face (see Young [2007] on the similar appeal o f "redeemed" old heads). By virtue of then reputations for tough- ness, their exploits in previous conflicts, and often their resources from the underground economy, these young adults sit atop a status system that defines masculinity and confers great cultural power, particularly in the eyes o f adolescent boys (Anderson 1999).7

7 One possible unintended consequence of inter- actions with older peers is that they may actually increase a boy's risk of victimization, as spending unstructured time with delinquent peers has the potential to expose boys to violent situations (Osgood et al. 1996). The boys in this study did not seem to recognize this risk. In addition, the level of interac- tion with older peers described here may not rise to the level required to increase this risk, as most of the

I briefly summarize here an example of how interactions with older peers structure boys' thinking in one seemingly unrelated domain, romantic relationships. For adolescent boys begirming to explore sexual and romantic rela- tionships, the stories and views of older peers play a role in structuring frames and expecta- tions regarding girls and girlfriends by defining the categories boys use to understand girls. These categories include "good girls" who are interested in real relationships and focus on school, "stunts" who are just after sex and may transmit an infection, and "golddiggers" who are after a man's money and may trick a boy into fathering a child. As a result, Roxbury Crossing and Franklin boys tend to approach potential girlfriends with hesitation, expecting them to try to take advantage of the relationship in some way. By contrast, these distinctions are far less salient for Lower Mills boys, and they approach relationships with more positive expectations, expecting most girls to be "good girls." The gender distrust that is created and legitimized i n part by interactions w i t h older peers i n Roxbury Crossing and Franklin has the poten- tial to affect relationship behavior and contra- ceptive practices among the boys from these areas (Harding forthcoming).

C O N C L U S I O N S

Theoretical perspectives on neighborhood effects, such as social isolation theory, assume that the neighborhood context serves as a source of socialization for youth. Through differential exposure to behavioral models or cultural ideas, disadvantaged neighborhoods are thought to influence how young people make decisions in domains such as schooling and romantic rela- tionships. Yet the empirical literature largely fails to identify and describe the sources and processes underlying such cultural transmis- sion, especially given recent research showing strong support for conventional cultural ideals among poor parents, community leaders, and other residents o f poor neighborhoods. Understanding cultural transmission processes in poor neighborhoods requires understanding with whom adolescents interact (compared with those in more advantaged neighborhoods), why

interactions occur within the relatively safe confines of the neighborhood.

VIOLENCE, ADOLESCENT BOYS, AND OLDER PEERS 461

iple o f how :ture boys' :d domain, scent boys lantic rela- Dlder peers :d expecta- ry defining tand girls, s" who are i focus on ;x and may rs" who are a boy into y Crossing a potential them to try ip i n some are far less y approach pectations, girls." The egitimized r peers i n the poten- nd contra- Tom these

hborhood ry, assume as a source lifferential tural ideas, thought to- ecisions in antic rela- rre largely jurces and transmis-

h showing ural ideals aders, and orhoods. .processes erstanding pared with nods), why

ife confines

those interactions occur, and their socialization consequences. Based on fieldwork with ado- lescent boys in three areas of Boston, this arti- cle proposes that disadvantaged older adolescents and young men are important sources of socialization for boys i n poor neigh- borhoods, thereby identifying one potential pathway for neighborhood socialization.

The NELS data indicate that adolescents in" disadvantaged neighborhoods are more likely than other adolescents to spend time with older individuals. A similar association is evident in the qualitative data, and the remainder of the analysis draws on these data to understand its causes and consequences. Greater interaction w i t h older peers among boys i n Roxbury Crossing and Fraruclin is in no small part a prod- uct of the level and organization of violence in these neighborhoods. When victimization is based on neighborhood identity, the salience o f such identities is magnified. Venturing outside one's neighborhood carries risk o f challenge, and youths from other neighborhoods are poten- tial enemies rather than potential friends. The neighborhood is one o f the only spaces where a young man feels safe from challenge from youths from rival neighborhoods, regardless of whether he is actively involved in neighbor- hood beefs. The result is a restricted set o f pos- sible friends, and older peers become a more attractive choice. Spending more time in the immediate neighborhood also increases boys' exposure to interactions with older neighbor- hood peers. Older peers become a source o f protection, through both then reputation and then capacity to intervene. By contrast, for the boys of Lower Mills, victimization is much less of a threat and these.coping strategies are large- ly unnecessary.

These strategies may have unintended con- sequences in other domains. Attachments to and interactions with older peers, however ben- eficial for safety and status, have the potential to expose young adolescent males to—and rein- force the legitimacy of—local frameworks regarding not only violence but also other domains, such as romantic relationships. Because the older peers available in Franklin and Roxbury Crossing often present models at odds with mainstream society, these interactions can influence the frames and strategies that boys bring to their decision making. The messages offered by the older peers are complex and at

times consistent with mainstream notions of responsibility, safe sex, and the importance of schooling. Nevertheless, the cross-cohort social- ization processes described here explain how adolescent boys encounter—and take serious- ly—local cultural models that can be at odds with mainstream models held by adults i n then communities or presented in the media or at school. In terms of cultural repertoires (Swidler 1986), this suggests that the repertoire elements selected for activation depend on social net- works and patterns of association.

The fieldwork on which this analysis is based is limited to adolescent boys, so it is not clear whether girls experience neighborhood vio- lence similarly. Boys have higher rates o f par- ticipation in violence, so they may be more affected by cross-neighborhood conflicts. Furthermore, conceptions o f masculinity drive responses to both individual and neighborhood- based challenges (Anderson 1999), and boys with absent fathers may be more susceptible to the influences of older male peers. Although gender differences in violence seem to be declin- ing (recent ethnographic research documents violent behavior among girls in inner-city neigh- borhoods), there is disagreement over the exis- tence o f gender differences i n strategies for dealing with violence (Cobbina et al. 2008; Jones 2004; Ness 2004). Moreover, some evi- dence suggests that violence among girls takes place more frequently in and around schools (Cobbina et al. 2008; Jones 2004).

In addition, the fieldwork and analysis focus on the experiences of urban African American and Latino boys in primarily African American neighborhoods in Boston, so it is not clear whether the cultural transmission processes described here operate in other neighborhoods or among youths from other racial and ethnic groups. Because the research design prioritizes cross-neighborhood comparisons, differences within neighborhoods by race and ethnicity are also outside the scope o f this study.

Moreover, it is possible that two features o f Boston neighborhoods led to the structure o f neighborhood-based conflicts and the small size o f neighborhoods described here: the relative- ly small size of Boston's public housing devel- opments, compared with those in Midwestern cities such as Chicago, and the citywide atten- dance areas for Boston high schools. In addition, compared w i t h cities like Chicago or Los

462 AMERICAN SOCIOLOGICAL REVIEW

Angeles, Boston has relatively little history o f race-based gang conflict or prominent "corpo- rate gangs" that control territory across large geographic areas. Only through further research can potential differences and similarities by gender, race, urban setting, and neighborhood be investigated and the generalizability of the findings assessed. Finally, while this study focuses on one particular strategy boys use to cope with violence, other strategies are also available and many boys employ multiple strate- gies. Future research should investigate how boys choose strategies and what explains vari- ation within neighborhoods.

This study elaborates some of the processes by which social isolation and social organiza- tion may account for neighborhood effects on adolescents. Social isolation theory must spec- ify how local cultural models regarding educa- tion, work, and relationships are transmitted within a neighborhood. I propose one cultural conduit: cross-age interactions between younger adolescents and older peers on the street. With respect to social organization theory, this arti- cle shows how the failure of a community to control violence can have spillover effects in other domains through the impact of violence on the age-structure of peer networks.

Another possible implication of these argu- ments is that older peers may play a role in other cross-generational processes, independent of the family-based intergenerational transmis- sion processes that are the focus of much strat- ification research. Suttles (1968) and Horowitz (1983) highlight the age segmentation of street- corner groups in inner-city neighborhoods. The argument here, while recognizing some degree of age segmentation, emphasizes that disad- vantaged neighborhoods, in comparison with more advantaged ones, have more cross-age social interaction. Such cross-age interactions may account for the transmission of social phe- nomenon such as norms about violence or frames regarding schooling.

More generally, these arguments suggest that violence is a critical characteristic of poor neigh- borhoods, structuring adolescents' daily lives and social networks. Considerable research examines the relationship between neighbor- hood disadvantage and crime and disorder (e.g., Sampson and Groves 1989; Sampson and Raudenbush 1999; Sampson et al. 1997), and psychologists focus on the cognitive and devel-

opmental effects o f exposure to violence (Aneshensel and Sucoff 1996; Margolin and Gordis 2000). However, few prior studies exam- ine the social effects of neighborhood violence. This article links violence, older peers, and socialization, suggesting that neighborhood vio- lence plays a role in the intergenerational trans- mission o f disadvantage.

David J. Harding is Assistant Professor in the Department of Sociology and Assistant Research Scientist at the Population Studies Center at the University of Michigan. His recent work includes "Collateral Consequences of Violence in Disadvantaged Neighborhoods " (Social Forces, June 2009), "Cultural Context, Sexual Behavior, and Romantic Relationships in Disadvantaged Neighborhoods" (ASR, June 2007) and Living the Drama: Community, Conflict, and Culture among Inner-City Boys, forthcoming from University of Chicago Press.

REFERENCES

Adler, Patricia A. and Peter Adler. 1998. Peer Power. New Brunswick, NJ: Rutgers University Press.

Akers, Ronald L . , Marvin D. Krohn, Lonn Lanza- Kaduce, and Marcia Radosevich. 1979. "Social Learning and Deviant Behavior: A Specific Test of a General Theory." American Sociological Review 44:636-55.

American Academy of Pediatrics. 2000. "Firearm- Related I n j u r i e s A f f e c t i n g the Pediatric Population." Pediatrics 105:888-95.

Anderson, Elijah. 1990. Streetwise. Chicago, I L : University of Chicago Press.

. 1999. Code of the Street. New York: Norton.

Aneshensel, Carol S. and Clea A. Sucoff. 1996. "The Neighborhood Context o f Adolescent Mental Health." Journal of Health and Social Behavior 37:293-310.

Benford, Robert D. and David A . Snow. 2000. "Framing Processes and Social Movements: A n Overview and Assessment." Annual Review of Sociology 26:611-39.

Black, Donald. 1993. The Social Structure of Right and Wrong. San Diego, CA: Academic Press.

Browning, Christopher R., Tama Leventhal, and Jeanne Brooks-Gunn. 2005. "Sexual Initiation in Early Adolescence: The Nexus of Parental and Community Control." American Sociological Review 70:758-78.

Centers for Disease C o n t r o l . 1997. "Rates o f Homicide, Suicide, and Firearm-Related Death among Children: 26 Industrialized Countries." Morbidity and Mortality Weekly Report 46:101-105.