Content uploaded by Callie H. Burt
Author content
All content in this area was uploaded by Callie H. Burt on Oct 23, 2017
Content may be subject to copyright.
Neighborhood Culture and Sexual Partnering
Neighborhood Social Processes and Adolescent
Sexual Partnering: A Multilevel Appraisal
ofAnderson’s Player Hypothesis
Mark T. Berg, University of Iowa
Callie H. Burt, University of Washington
Man-Kit Lei, University of Georgia
Leslie Gordon Simons, University of Georgia
Eric A. Stewart, Florida State University
Ronald Simons, University of Georgia
An increasing amount of sociological research is directed at unpacking the social
processes behind neighborhood effects on youth behavioral outcomes. The goal
of the current study is to build upon these prior efforts and advance research on
neighborhood cultural mechanisms and adolescent sexual-partnering behaviors. We
formulate and test a series of multilevel hypotheses informed by Anderson’s research
on the relationship between neighborhood cultural processes and multiple sexual part-
nering or “player-like” sexual behavior. The hypotheses are tested using data from a
multisite sample of African American adolescent males from the Family and Community
Health Study. Results from a series of multilevel models provide evidence of both con-
textual and indirect effects of neighborhood culture on sexual partnering; moreover,
these effects are conditioned by neighborhood structural deprivation.
For decades, sociologists have recognized that neighborhoods operate as an
important social context for adolescent development. Adolescence is a period of
increasing participation outside the home, and it is during this time in which
neighborhoods exert potentially powerful effects on romantic relationships, fer-
tility outcomes, and contraceptive use (see Giordano 2003). More recently,
researchers have increasingly focused on identifying how neighborhoods come to
affect variation in youth outcomes, including sexual activities (see Sampson
2012). Although there is mounting empirical evidence that dimensions of neigh-
borhood social organization inuence adolescent sexual outcomes, the role of
spatially embedded cultural mechanisms is not well understood. Relatively few
studies specify cultural measures and, in some cases, researchers must instead rely
on proxy indicators of these constructs. However, recent research suggests that
adolescent sexual activities vary with the cultural context of local neighborhoods,
Address correspondence to Mark T. Berg, University of Iowa, Department of Sociology, W126 Seashore
Hall, Iowa City, IA 52242; Phone: 319-335-2495; Email: mark-berg@uiowa.edu.
Neighborhood Culture and Sexual Partnering 1823
© The Author 2016. Published by Oxford University Press on behalf of the
University of North Carolina at Chapel Hill. All rights reserved. For permissions,
please e-mail: journals.permissions@oup.com.
Social Forces 94(4) 1823–1846, June 2016
doi: 10.1093/sf/sow032
Advance Access publication on 00 Month 2016
albeit in potentially complex ways (e.g., Harding 2010). Establishing a strong
empirical understanding of how cultural processes operate is critical to advancing
a process-oriented framework of how neighborhood context brings about varia-
tion in sexual behaviors.
Among the more inuential conceptual frameworks, Elijah Anderson’s (1989,
1999) perspective offers an explicit set of assumptions regarding cultural pro-
cesses and sexual-partnering practices among young men (see Giordano etal.
2009). According to Anderson (1989), the social context of poor neighborhoods
emphasizes a “player orientation” that deems sexual encounters with multiple
women as an “important marker of social status” (60). By his account, the “player
culture” operates as an emergent social property of neighborhoods that affects
sexual encounters whether or not young men personally subscribe to its rules.
Asmall number of important studies have examined Anderson’s assumptions
providing some support; yet, the multilevel interpretations implicit to his perspec-
tive remain theoretically underdeveloped (see Giordano etal. 2009; Warner etal.
2011). Indeed, existing research has considered neither the contextual vis-à-vis
individual dimensions of “player” cultural processes nor the potential multiplica-
tive effects of structural deprivation. With the goal of contributing to these
knowledge gaps, the present paper formulates and tests a multilevel perspective
on neighborhood cultural processes and sexual partnering, most notably the
assumptions implicit to Anderson’s perspective.
Altogether, the paper pursues two main goals: First, we examine hypotheses
about the mechanisms through which the neighborhood player culture purport-
edly affects sexual partnering among males. According to one hypothesis, the
cultural messages embodied in the “player culture” inuence multiple sexual
partnering despite adolescents’ personal evaluations of this behavior. Hence, ado-
lescents allegedly engage in player-like sexual activities to satisfy local status con-
cerns and avoid negative peer sanctions, even if doing so contradicts their own
cultural frames. If this “contextual” proposition is operative, neighborhood
“player” culture should predict sexual partnering outcomes regardless of individ-
ual-level frames. An alternative hypothesis, however, posits that the neighbor-
hood player culture operates indirectly through its inuence on the personal
salience of player frames. Where this culture is prominent in the landscape, young
men incorporate its logic as part of their sexual decision-making frame. Put in
practical terms, individual-level partnering frames should mediate the effects of
the neighborhood culture. Second, we examine the nexus of poverty and cultural
processes by considering whether the effects of partnering frames vary across
levels of neighborhood disadvantage. Specically, we propose a new “status
deprivation hypothesis” that posits that the accumulation of multiple sex part-
ners provides an alternative pathway of status attainment for young men in low-
income neighborhoods. Under this assumption, structural deprivation is
hypothesized to have a magnifying effect on the relationship between player
frames and the accumulation of sex partners.
We exa mine th ese pr edic tion s w ith or igina l data fr om a mult isite sample of ado-
lescent African American males. Taken together, this research builds on prior efforts
to contribute to the theoretical literature on neighborhood effects and adolescent
sexual conduct by casting additional light on the implications of local cultural
1824 Social Forces 94(4)
processes. Furthermore, understanding the social mechanisms related to sexual
decision-making is an issue of signicant public health importance. For instance,
young people who have numerous sex partners outside stable romantic relation-
ships are at higher risk for sexual assault and negative health outcomes, including
sexually transmitted diseases (see McCarthy and Grodsky 2011). Also, this behav-
ior increases the likelihood of experiencing multiple-partner fertility, which con-
tributes to intergenerational inequality (see Carlson and Furstenberg 2006).
Background
Theorizing Neighborhood Social Processes and Adolescent Outcomes
Adolescents are particularly susceptible to neighborhood processes because their
mobility tends to be geographically limited; furthermore, it is at this stage of
development in which they are particularly inuenced by the behaviors of their
same-sex peer groups (Giordano 2003). Neighborhoods are thought to provide
contextual mechanisms involving collective regulation and socialization (see
Berg, Sevell, and Stewart 2015; Leventhal and Brooks-Gunn 2000). According to
theories of neighborhood effects, socially disorganized communities are dened
by (1) a collective breakdown in the capacity of the primary group to control the
conduct of its members; and (2) the existence of competing cultural systems that
sanction unconventional behavior (Berg etal. 2013; Kornhauser 1978; Shaw and
McKay 1969, 320). Contemporary sociological models mainly emphasize neigh-
borhood regulatory or informal control dynamics such as collective efcacy
(Sampson 2012, 150–56). According to these social control perspectives, neigh-
borhoods vary in the extent to which they “provide opportunities for and con-
strain sexual interactions” among adolescents (Browning and Olinger-Wilbon
2003, 732; Browning, Leventhal, and Brooks-Gunn 2005).
Neighborhood Disadvantage, Culture, and Sexual Partnering
Although neighborhood-effects research has focused primarily on control or
regulatory mechanisms, this work has largely neglected the potential role of cul-
ture (Small, Harding, and Lamont 2010). As Sampson and Wilson (1995, 53)
note, scholars “dismissed the relevance of culture based on analysis of census data
that provide no measures of culture whatsoever.” However, Wilson’s (1996)
research on urban poverty reignited scholarly interest in neighborhood culture as
an explanatory mechanism. Within this vein, Sampson and Wilson (1995) pro-
posed that the ecological concentration of inequality fosters “cognitive land-
scapes” or ecologically structured norms about appropriate conduct that tend to
sanction risky and unconventional conduct. Modern research on urban inequal-
ity now generally agrees that blocked access to conventional opportunities gives
rise to a cultural emphasis on alternative pathways of status attainment—an
emphasis that is typically not as acute in middle-class neighborhoods, owing to
the limited availability of conventional pathways (e.g., Crane 1991; Jencks and
Mayer 1990; Massey and Denton 1993). This line of research species the con-
vergence of cultural processes and inequality as sources of risky sexual conduct.
For low-income youth, the “opportunity costs” for fathering children out of
Neighborhood Culture and Sexual Partnering 1825
wedlock are lower given that “the occasions for economic and social advance-
ment are so minimal” (Furstenberg etal. 1987, 512). Likewise, ethnographers
observe that youth from poor neighborhoods are more apt to “turn to romantic
and sexual relationships as a way to secure material resources and social status”
(Harding 2010, 200). For instance, a study found that the challenges of material
deprivation led many young women to forego more conventional routes to family
formation (Edin and Kefalas 2005; see Molborn 2010).
Anderson’s Description of the Player Culture
Among recent theoretical developments on neighborhood mechanisms, Elijah
Anderson’s (1989, 1999) account expands the discussion of cultural processes
relating to adolescent sexual encounters. His research describes the content of
sexual-partnering norms in disadvantaged neighborhoods. According to Anderson
(1999), poor urban areas harbor a “street culture” that emphasizes certain styles
of interpersonal behaviors typically deemed unacceptable to wider society.
Aprominent feature of the street culture is a complex of frames or shared under-
standings related to promiscuous sexual conduct, known otherwise as the “player
culture.” Anderson observes that many young men in these settings emphasize
sexual prowess as a dening feature of manhood. They purportedly boast of their
own sexual exploits during the course of peer interactions, and position them-
selves as accomplished “players” in the sexual game in which the primary goal is
to have as many sex partners as possible. For these young men, sex may be casual
in its enactment, but it is taken as a serious measure of a young man’s worth. As
Anderson (1989, 60) notes, “to many boys sex is an important symbol of local
status; sexual conquests become so many notches on one’s belt.” A woman is
treated by some young men as a sexual object—a prize to “be won for his per-
sonal aggrandizement” (Anderson 1999, 150). Qualitative studies also reveal that
boys from poor neighborhoods view the player role as instrumental to demon-
strating their masculinity (Lopez 2014). To be sure, peer groups in middle-class
neighborhoods also espouse the messages of the player culture; yet, as Anderson
describes, because youth in disadvantaged places have comparatively fewer
opportunities to earn status in conventional arenas, the “player” lifestyle is given
more emphasis as an alternative method of status attainment.
Anderson’s (1999) perspective suggests that the player culture is not necessar-
ily dominant to the exclusion of conventional cultural models of sexual-
partnering strategies. Rather than being a monolithic entity, the player culture
coexists alongside a “decent orientation”; the latter denes acceptable sexual
behavior according to the standards of conventional society (i.e., sex reserved
primarily for long-term, monogamous relationships). Decent families instill in
their children the tools needed to achieve success in conventional avenues, while
teaching them to avoid risky behaviors that could jeopardize their goals. Although
adherents of decent orientations are a numerical majority—even in poor neigh-
borhoods—the player culture still holds considerable sway due to the powerful
status markers it confers. Specically, Anderson (1999) observes that the player
culture weighs heavily on the sexual decision-making of young men even if they
do not personally subscribe to its rules (Anderson 1989). This logic implies that
1826 Social Forces 94(4)
the player culture exerts a contextual effect. For example, although some young
men may be genuinely interested in a long-term relationship and possess roman-
tic feelings for their female partner, they are reluctant to convey this sentiment
whether publicly or privately because it would clash with the player script.
According to Anderson (1999), the social inuence of the player culture is illumi-
nated during occasions when young men allow a woman to “control him” in an
exclusive relationship. When this occurs, the group may assign crude and demean-
ing labels to the young man, pushing him to the very margins of their social circle.
Recent Interpretations of Neighborhood Cultural Dynamics
The logic of Anderson’s (1999) reasoning, like that of other neighborhood-effects
perspectives (e.g., Wilson 1996), appears to reect a conceptual model in which
the local culture provides the values that distinctively motivate action. However,
analysts have critiqued the conceptual assumptions of these “values-based”
cultural frameworks for suggesting that individuals enact cultural rules “unprob-
lematically” (DiMaggio 1997, 265; Small and Newman 2001). For instance, per-
spectives such as these also struggle to explain why people often depart from their
own values to engage in behaviors they publicly condemn (Berg, Sevell, and Stewart
2015). An alternative “relational” conceptualization of cultural effects argues
that culture is a set of scripts or “sense-making” tools that people selectively
deploy to guide their actions (see Hays 1994; Vaisey 2008). Culture therefore
inuences action not only by providing the values to which “action is oriented,
but by shaping a repertoire or ‘tool kit’ of habits people use to construct ‘strate-
gies of action’” (Swidler 1986, 273). Similarly, cultural frames are part of an
individual’s tool kit, serving as lters through which they understand the way the
world operates (Benford and Snow 2000). Frames dene the expectations about
the practical consequences of a set of actions (Goffman 1974; see Hannerz 1969),
but remain sensitive to the burdens arising from local structural constraints.
Researchers have recently modied the assumptions about the cultural land-
scape of urban neighborhoods based on the conceptual logic of the cultural script
framework. For instance, Small’s (2004) study of a Boston neighborhood demon-
strated how cross-cohort differences in cultural frames, and not necessarily values,
accounted for the relative inaction of a younger cohort to engage local neighbor-
hood problems. According to Harding’s (2007) research, impoverished neighbor-
hoods foster a mélange of competing standards about the appropriateness of
sexual activity. His work suggests that young men often balance oppositional cul-
tural messages with conventional standards, which consequently complicates their
sexual decision-making. Much like the assumption of Anderson’s (1999) model,
strands of this research generally agree on two notions: (1) that poor neighbor-
hoods harbor a diverse climate of cultural messages (e.g., decent and street); and
(2) that the degree of social support granted to a certain behavior—or its cultural
salience—strongly inuences the decisions young people make (see Berg etal.
2013). For instance, as Harding (2010) writes, “when others worthy of respect
adopt a cultural model, an adolescent evaluating his options will be more likely to
follow suit” (147). Across these analytical accounts, the neighborhood culture is a
“constraining and enabling” source of decision-making (Hays 1994, 65).
Neighborhood Culture and Sexual Partnering 1827
Taken together, multiple strands of research assume that the local cultural
context has important implications for adolescent sexual activities, and its effect
is amplied in poor environments. Moreover, this work underscores the notion
that, even where elements of the player culture are a numerical minority, this
orientation inuences how young men approach sexual relationships.
Prior Empirical Research
A small number of studies have tested hypotheses about the effects of cultural
and other ecological processes on adolescent sexual activities, some of which do
so with Census-based socioeconomic variables specied as proxy measures of
these constructs. Altogether, ndings from this line of research paint an inconsis-
tent picture of the direct effects of neighborhood disadvantage, whether it is
treated as a proxy measure or as a control variable. For example, studies show
that dimensions of neighborhood poverty are not directly related to the number
of sex partners or the prevalence of sexual activity among adolescents (e.g.,
Browning and Olinger-Wilbon 2003; Giordano etal. 2009; Dupere etal. 2008;
Roche etal. 2005), while other research observes positive effects (Baumer and
South 2001; Brewster, Billy, and Grady 1993). Although this body of research is
inuential, it is somewhat difcult to draw conclusions from it about the impact
of player-specic cultural processes on sexual activities because they are not
directly measured (Lauritsen 1994). Proxy measures can capture contradictory
mechanisms (e.g., social control, institutional, and cultural mechanisms), none of
which can be readily detected from the estimates.1
Also, Harding’s (2007) research assessed whether collective agreement about the
appropriateness of pregnancy at the neighborhood level (i.e., cultural heterogene-
ity) predicted changes in the prevalence of adolescent sexual behavior. Although his
study provides signicant insights regarding neighborhood cultural mechanisms, it
does not speak directly to the empirical validity of Anderson’s perspective since it
is not an explicit test of hypotheses regarding the cultural basis—particularly the
status value—of player-like sexual practices. Also, a small body of research focuses
explicitly on the context of multiple partnering using constructs assessing neighbor-
hood- or individual-level cultural constructs measuring player-like cultural mecha-
nisms. For instance, Warner, Giordano, Manning and Longmore (2011) found that
the number of lifetime sexual partners is a function of a sexually permissive neigh-
borhood normative climate, which lends some empirical support to Anderson’s
assumptions. Consistent with Anderson’s (1999) arguments, Giordano etal. (2009)
also discovered that youth who subscribed to the player identity were likely to have
a greater number of sexual partners. Also, their study reported that some of the
young men who exhibited player-like sexual behavior did not necessarily endorse
the player culture; instead, their behavior was an attempt to forge a positive iden-
tity among their peers, which is consistent with a contextual effect.
While existing work has cast important light on the etiology of adolescent
sexual-partnering practices, researchers have yet to simultaneously examine the
inuence of a neighborhood player culture and individual-level frames on mul-
tiple sexual partnering. As a result, it remains unclear whether youth engage in
player-like sexual behavior out of concerns to project a favorable identity or
1828 Social Forces 94(4)
because they personally endorse player norms. Furthermore, prior research has
not considered whether the effects of personal adherence to the player orientation
on sexual behavior are stronger in structurally deprived neighborhoods—a theme
implied throughout this body of literature.
Current Study
Building on prior research, the purposes of this study are twofold. First, we take
a process-oriented approach to neighborhood effects by examining the mecha-
nisms through which the neighborhood player culture operates to inuence sex-
ual-partnering behaviors. We situate our conceptual framework in the logic of
cognitive perspectives from the sociology of culture (e.g., Harding 2010; Small,
Harding, and Lamont 2010). Figure1 illustrates our assumptions. On one hand,
perhaps neighborhood cultural processes have a direct contextual inuence on
sexual-partnering behavior as a result of a compliance mechanism. According to
this hypothesis, neighborhood player culture should increase multiple sexual
partnering irrespective of the salience of player frames to the individual.
AsAnderson (1999) notes, there are status rewards for conforming to the rules
Figure1. Conceptual model of the study expectations
Neighborhood
Partnering Cultural
Climate
Individual-level
Partnering Frames
A
B
AB
C
Individual-level
Partnering Frames
Multiple Sexual
Partners
Structural
Deprivation
Multiple Sexual
Partners
Neighborhood Culture and Sexual Partnering 1829
of the player culture, and there are also penalties for failing to do so. The C path
in panel A of gure 1 represents a compliance effect whereby the neighborhood
cultural climate inuences sexual behavior independent of individual frames. Evi-
dence of this would point to the player culture as an emergent social property—a
property that inuences sexual conduct irrespective of individual characteristics.
An alternative hypothesis predicts that the neighborhood player culture indi-
rectly inuences sexual-partnering behavior through its effects on individual
player frames. Under this assumption, youth come to adopt the elements of the
player culture as decision-making frames, and therefore their behavior reects the
personal salience of the frame as opposed to mere conformity to the neighbor-
hood culture (see Cialdini and Goldstein 2004). If this is the case, the neighbor-
hood player culture should have indirect effects on multiple partnering through
individual-level partnering frames. This mediation process is represented by the
combined effects of the A and B paths in gure 1, panel A. Also, it is possible that
both aforementioned analytical processes channel the total effects of neighbor-
hood player culture.
Second, we examine a “status-deprivation” hypothesis implicit to existing
research that suggests the “player” culture is more salient to adolescents in struc-
turally disadvantaged neighborhoods. As described earlier, the theoretical basis
for this claim assumes that sexual experiences with multiple women are an alter-
native mechanism of status attainment for youth located in places with limited
opportunities for conventional advancement. A status-deprivation prediction
therefore holds that structural disadvantage magnies the effects of individual-
level adherence to player frames on multiple sexual partnering. This hypothesis
is illustrated in panel B of gure 1, where a pathway from structural deprivation
affects the link between individual frames and sexual partnering.
Data and Methodology
Sampling Strategy
This study is based on data from the Family and Community Health Study
(FACHS), a multisite investigation of neighborhood and family effects on health
and development among African American adolescents (see Simons etal. 2011).
The rst wave of data was collected in 1997, wave 2 in 1999, wave 3 in 2001, and
wave 4 in 2003. Respondents were, on average, 10.5, 12.5, 15.5, and 18.5 years
old at each wave. At wave 1, the participants included 867 children (400 boys
and 467 girls) and their primary caregivers. In the current study, we use waves 3
and 4, in which data are available on the social context of sexual behaviors. As
discussed elsewhere (Simons etal. 2011), there was little evidence of selective
attrition across the four study waves. The sample for our study is composed of
312 males who had complete data at wave 4; data on 88 (or 22 percent) of the
boys were missing due to sample attrition. This attrition rate is comparable to
other longitudinal multisite studies. We focus solely on young males since the
literature from which the hypotheses are derived concerns the conduct of this
group. Also, there are key gender differences in beliefs and involvement in casual
sexual activities (e.g., Cubbins and Tanfer 2000).
1830 Social Forces 94(4)
During the initial wave of data collection, Census block groups (BGs) were
used to initially identify potential sampling neighborhoods. Using 1990 Census
data, the FACHS team selected BGs in both Iowa and Georgia in which the per-
cent of African American families was high enough to make recruitment eco-
nomically practical (10 percent or higher), and in which the percent of families
with children living below the poverty line exceeded 10 percent. Using these cri-
teria, 259 BGs were identied (115 in Georgia and 144 in Iowa). The families
from these block groups were contacted; if they consented, they were selected for
recruitment. In Georgia, families were recruited from metropolitan Atlanta areas.
In Iowa, all BGs were located in Waterloo and Des Moines. Note that Census
block groups were used in wave 1 only as a sampling frame, but in all later waves
the physical addresses for each respondent were used to obtain their Census tract
residence. Using Dalaker’s (2001) criteria from the 2000 Census, approximately
15 percent of FACHS families resided in rural settings and the remainder were
located in metropolitan areas.
Current Study Measures
At wave 3, respondents were distributed across 280 Census tracts; since the
majority of these tracts contained fewer than ve respondents, it is methodologi-
cally challenging to create reliable neighborhood cultural measures; further, esti-
mates from multilevel regression models are unlikely to be reliable. Similar to
prior research, we implement a clustering procedure to group Census tracts
together into larger “synthetic” neighborhoods that are internally homogeneous
on several theoretically relevant socioeconomic characteristics (see Sampson,
Raudenbush, and Earls 1997; Clarke and Wheaton 2007). Wefollowed the pro-
cedures of prior sociological research to develop the clusters or synthetic neigh-
borhoods, particularly the program of research by Sampson etal. (1997). For the
initial step, we obtained information on the 280 Census tracts in Iowa and Geor-
gia where respondents were located based on their physical addresses reported at
wave 3. Next, we used the “Fossil” R-Package to obtain the geographic distances
of the 280 Census tracts from one another—tracts containing respondents’
addresses—using longitude and latitude coordinates with the goal of identifying
geographically proximate areas. Next, from this strategy we identied 14 geo-
graphically proximate areas within which the 280 Census tracts were located
(7 = Iowa, 7 = Georgia). Following prior research (e.g., Sampson, Raudenbush,
and Earls 1997), we then sought to create clusters that were internally compa-
rable on the following 2000 six Census indicators: per capita income, proportion
of households on public assistance, percentage of impoverished households, per-
centage of unemployed adults, percentage of female-headed families, and per-
centage of non-Hispanic blacks. Similar to prior work, these six items loaded on
a single factor in a principal components analysis (Sampson, Raudenbush, and
Earls 1997). Next, we used Ward’s minimum variance method in the SAS Cluster
program to perform analyses on the single socioeconomic factor to place the
tracts into homogeneous neighborhood clusters within each geographic region.
Altogether, the process generated 50 neighborhood clusters or synthetic neigh-
borhoods: 20 situated in Georgia and 30 in Iowa. The Census tracts within a
Neighborhood Culture and Sexual Partnering 1831
cluster are internally homogeneous, spatially proximate, and share a common set
of socioeconomic characteristics. Families allocated in a particular neighborhood
cluster were considered to be living within similar ecological contexts (see Sampson
2012). Clarke and Wheaton (2007) found that cluster strategies attenuate the
overestimation of group-level variance components, and increase model perfor-
mance, including reliability estimates.
A well-known challenge in neighborhood-effects research is that people do not
randomly select into environments (Sampson 2012). Since our research does not
randomly allocate participants to neighborhoods or use an instrumental variable
procedure, we are unable to identify a purely unbiased estimate of the causal role
of player culture; the models essentially estimate a conditional correlation. How-
ever, we employ multiple theoretically relevant controls in an attempt to control
for potential observed confounding mechanisms.
Dependent Variable
Table1 displays descriptive information for each of the study variables. Data for
the dependent variable, the number of lifetime sex partners, is based on a measure
from wave 4 that asks respondents: “With how many different people have you
had sex?” Responses are assessed on a ve-level ordinal scale: (0) None, (1) One
or two, (2) Three or four, (3) Five or six, (4) Seven or more. Approximately 13.4
percent of males reported having no sexual partners, whereas 16 percent reported
one or two, about 27.5 percent reported three or four, 14.1 percent reported ve
or six sex partners, and more than 28.8 percent of the sample had more than
seven partners.
Independent and Control Variables
Individual-level variables The primary independent variable at the individual
level, labeled partnering frames, is a summed measure of respondents’ support for
the player culture based on a set of ve items regarding the social acceptability of
multiple sexual partners. The items were developed by FACHS investigators to
assess adolescent-health behaviors (see Gibbons, Gerard, and Lane 2003). FACHS
respondents read an introductory statement: “We’re not thinking about anyone
in particular, just your image of people your age who ‘sleep around’ (have mul-
tiple sex partners).” Then they are asked ve questions: “The type of [BOY] about
your age who ‘sleeps around’ (has multiple sex partners): (1) How popular are
they? (2) How smart are they? (3) How cool are they? (4) How attractive are
they? (5) How dull or boring are they (reverse coded)?” The four potential
responses ranged from (1) “not at all” to (4) “very.” Each item refers to the inten-
sity of status appraisals assigned to other males who have had sexual intercourse
with multiple women; the degree of status assigned is indicative of the respect
granted to individuals who have numerous partners. As such, the items capture
the conceptual components explicit to the theoretical arguments put forth by
Anderson (1999) and others (Giordano etal. 2009) and are consistent with prior
measurement strategies to index individual-level cultural mechanisms (Harding
2007). Given the conceptual foundation, this operationalization assumes that
1832 Social Forces 94(4)
frames do not reect xed commitments to styles of actions but reect a super-
cial adherence to cultural classication (see Vaisey 2008, 607–8). Altogether, the
summed scale exhibits rather strong internal reliability (alpha = .85).
A series of theoretically relevant control variables were added to the models;
when available, we incorporated measures from wave 4, and elsewhere the vari-
ables were taken from wave 3. Certain variables are not available across all
FACHS waves. To control for the potential confounding effects of risky tenden-
cies, we include a variable assessing antisocial behavior at wave 4. This variable
is measured using a variety count of youth self-reports on the conduct disorder
section of the Diagnostic Interview Schedule for Children (DISC-IV) (American
Psychological Association 1994). The section contains questions regarding
prior-year commission of 26 antisocial acts (e.g., shoplifting, physical assault,
vandalism).
We also include an item from wave 4 capturing respondents’ goals to achieve
conventional markers of life-course attainment, otherwise known as success
Table1. Study Variables Descriptive Statistics
Dependent Variable
Number of Prior Sex Partners n %
(0) None 42 13.4
(1) One or two 50 16.0
(2) Three or Four 86 27.5
(3) Five or Six 44 14.1
(4) Seven or more 90 28.8
Mean SD Min Max
Neighborhood Level
Partnering cultural climate 1.31 .25 .20 2.4
Collective efcacy −.045 .31 −1.46 .53
Structural deprivation .069 .89 −1.47 2.67
Sex ratio .797 .11 .55 1.08
Individual-Level
Partnering frames 1.53 .71 0 4
Antisocial behavior 3.37 4.02 0 21
Success goals 1.41 .66 1 5
Family pregnancy reaction 2.07 1.42 1 8
Pregnancy impact 2.37 1.47 1 9
Family structure 1.58 .49 1 2
Authoritative parenting 3.27 .369 1.44 4
Steady relationship .332 .471 0 1
Annual income 5.56 9.74 0 99
Age (Months) 231 9.71 202 257
Neighborhood Culture and Sexual Partnering 1833
goals. Youth with aspirations for success are less likely to engage in behaviors
that could jeopardize their future (Halpern etal. 2000), and therefore controlling
for success goals is important to manage threats of omitted-variable bias. Youth
were asked the following question: “How important is it for you to be successful
at work or your career?” (1 = “extremely” to 4 = “not very important”). Also,
prior research suggests that the consequences of impregnating a women and later
caring for a child are a deterrent to risky sexual behavior. We assess the pregnancy
impact at wave 3 with a question that asks respondents “how they would feel if
they got a girl pregnant.” Respondents range from (1) “I would be very upset” to
(5) “I would be very pleased.”
Parenting practices involving monitoring and social support predict involve-
ment in risky behaviors (see Manning, Longmore, and Giordano 2005); it is
important to control for these processes since they may explain any effect of
cultural processes on sexual activities. A family structure variable captures
whether there is a second caregiver in the home of the respondent, which reects
parental supervision capacity. Next, we include a summary measure of authorita-
tive parenting from wave 3. Authoritative parents typically hold strict rules
regarding their child’s social activities and espouse high expectations for their
conduct. This measure consists of combined child and caregiver reports from
wave 3 regarding various key parenting practices (see Simons etal. 2005). Higher
scores indicate better-quality parenting. Parents’ reactions to involvement in sex-
ual behavior, including fertility experiences, can affect whether youth engage in
sexual activities (Halpern etal. 2000). To account for these effects, we include an
ordinal measure assessing family pregnancy reaction from wave 4, which asks:
“How would adults in your family react if you got a girl pregnant now?” Respon-
dents chose from ve potential responses: (1) “they would be very upset” to (5)
“they would be very pleased.”
We incorporate a set of demographic variables from wave 4, including current
relationship status or single: (1) single; or (0) dating a partner, living with a part-
ner, or married. Adolescents in a steady relationship are more committed to sex-
ual exclusivity (Paik and Woodley 2012). A measure of respondents’ age
(inmonths) was added as a control variable. Also, we control for their caretakers’
annual income with an ordinal measure indexing how much income they received
in the past year; there are 17 categories ranging from (0) none through
(17)$200,000 or more.
Neighborhood-level variables Similar to prior research, the neighborhood-
cluster-level variables are aggregated at the Census tract to the cluster level (see
Sampson, Raudenbush, and Earls 1997); specically, both the Census-based mea-
sures and the survey-derived measures are aggregated from the multiple tracts
within a cluster and then averaged to create a neighborhood-cluster-level vari-
able. A measure of the partnering cultural climate is the key independent variable
at the neighborhood level. Similar to the designs of prior research, the measure is
derived by aggregating individual composite scores to the neighborhood-cluster
level and calculating the mean of the individual scores on the player frames
variable (see Bernburg and Thorlindson 2005). A multi-item composite measure
reduces measurement error. The variable is mean-centered in the analysis to iso-
late its contextual and compositional components. Specifying neighborhood- and
1834 Social Forces 94(4)
individual-level measures of the cultural construct addresses the possibility that
in the absence of one measure the other might operate partly as a proxy for
effects operating at another level.
A measure of structural deprivation is constructed from six tract variables
gathered in the 2000 Census: per capita income, proportion of households on
public assistance, percentage of impoverished households, percentage of unem-
ployed adults, percentage of female-headed families, and percentage of non-
Hispanic blacks. A factor analysis of these variables extracted a single factor with
an Eigenvalue exceeding 2.5 from which regression-based factor scores were
computed. As noted earlier, this deprivation measure was averaged across tracts
to create a neighborhood-cluster-level score. We also include a measure of collec-
tive efcacy that was assessed with two subscales at wave 4—one measuring
social cohesion and the other social control; this variable was adopted from the
PHDCN (see Sampson, Raudenbush, and Earls 1997). The social control items
ask respondents to report how likely people in their communities will intervene
in eight problematic behaviors; and the cohesion construct refers to 15 items
reecting pro-social interactions among the respondents’ neighbors. The two sub-
scales were standardized and averaged to form a composite indicator (e.g.,
Simons etal. 2005). A measure of the ratio of unmarried males to females age 15
and older or the sex ratio is available in the 2000 Census. Prior work suggests
that the availability of unmarried females in local areas has implications for part-
nering behavior (Warner, Manning, etal. 2011) and other sexual conduct (Green
etal. 2012). It is worth noting that high levels of incarceration can exert an exog-
enous inuence on the sex ratio.
Analytical Plan
To e xa min e h ow n ei ghbo rh oo d- a nd i nd iv id ual -l ev el v ar ia ble s a re r el ated to m ul -
tiple sexual partnering, we use a multilevel ordered logistic model (MOLM),
which is an extension of a random intercept binary logistic regression model. The
general functional form of the MOLM is as follows: Suppose a model includes
Xij, the independent variable, with Zij being the neighborhood-level moderator
variable, then the model for the cumulative response probability for individual i
in neighborhood j is
logPr()
Pr() ,
yk
yk XZXZ uu
ij
ij kijijijijoj
≤
⎛
⎝
⎜⎞
⎠
⎟=+ ++ +
>αβ ββ
1230j
ju
N~(
,)
,0 0
2
σ
where response variable y takes values k = 1, . . . C-1, log(Pr()Pr
()
)yk
yk
ij ij
≤
/
>
represents the cumulative logit of the outcome for individual i in neighborhood j,
α
k represents threshold parameters of ordinal outcome k or multiple sexual part-
nering,
β
1 and
β
2 are the xed effects of X
ij
and Z
ij
,
β3
is the coefcient of the
product term of Xij Zij , and uoj is a random intercept.2
Also, we employ MOLM with a structural path model to investigate the indi-
rect and direct effects of neighborhood player culture and individual player
frames. We estimate the condence intervals around the point estimates of the
Neighborhood Culture and Sexual Partnering 1835
mediating pathways (see Krull and MacKinnon 2001). Per the study hypotheses,
we specify a 2–1–1 multilevel mediation framework whereby the effects of neigh-
borhood player culture (level 2) operate indirectly through individual player
frames (level 1) on sexual partnering (level 1) (see Preacher, Zyphur, and Zhang
2010). Finally, we examine the cross-level interaction (
Xij
Z
ij) between neighbor-
hood structural disadvantage and individual player frames to assess the moderat-
ing effects assumption. The MOLM and multiplicative models are estimated in
MPLUS 7.01.
Results
Table2 displays the estimates from a series of three MOLM models. The rst
model contains the full range of study variables except the neighborhood partner-
ing climate and the individual partnering frames variables. According to the esti-
mates, neither neighborhood socioeconomic characteristics nor collective control
mechanisms are signicantly related to sexual partnering. This pattern is consis-
tent with some prior research that nds that concentrated disadvantage does not
have additive effects on short-term sexual partnerships (see Warner, Giordano,
Manning and Longmore 2011); also, prior work has found that collective ef-
cacy exerts multiplicative rather than additive effects (e.g., Browning and Olinger-
Wilbon 2003). Turning now to the individual-level estimates, the results from
model 1 show that a unit increase in antisocial behavior (b = .123, SE = .023) is
related to an increase in the log-odds of being in a higher sexual-partnering cat-
egory. Looking at the family structure variable, it appears that boys from single-
parent homes are at greater risk for being in higher categories of sexual partnering.
Also, estimates from the age variable indicate that older respondents are likely to
report having more sexual partners.
Turning now to the initial set of study hypotheses, model 2 introduces the
player normative climate variable at the neighborhood level alongside an array
of control variables. Consistent with Anderson’s (1999) notions about contextual
effects, the estimates show that a one-unit increase in the player cultural climate
increases the ordered log-odds of being in a higher sexual-partnering category
(b = 1.52; SE = .521). This pattern of results is consistent with theoretical expec-
tations: Youth who are exposed to a neighborhood cultural climate favoring
“player-like” behaviors are more likely to have a higher number of sex partners.
Estimates for the remaining variables remain relatively stable across the two
models, demonstrating modest effect-size differences.
Next, model 3 incorporates individual-level player frames into the equation.
This model allows us to directly investigate whether the neighborhood player
climate operates to inuence multiple sexual partnering either as a contextual
effect, or indirectly as a mediation process. Recall that strong evidence of the
former would emerge if the neighborhood partnering climate variable has sig-
nicant effects independent of the individual-level frames measure. Looking at
model 3, the results indicate that multiple sexual partnering is a function of both
the neighborhood cultural climate (b = 1.16, SE = .538) and the individual-level
player frames (b = .636, SE = .196). Based on these estimates, it appears that both
1836 Social Forces 94(4)
Table2. Multi-level Ordered Logistic Regression Model Predicting Sexual Partners
Model 1 Model 2 Model 3 Model 4
Coef. S.E. Coef. S.E. Coef. S.E. Coef. S.E.
Neighborhood Level
Partnering cultural climate – – 1.52* .521 1.16* .538 1.21* .542
Collective efcacy −.285 .475 −.021 .514 −.121 .520 −.210 .542
Structural deprivation −.133 .178 −.027 .173 −.028 .176 −.522 .276
Sex ratio −.737 1.32 −1.17 1.31 −1.26 1.175 −1.69 1.27
Cross-level
Structural deprivation × partnering frames – – – – – – .315* .149
Individual-Level
Partnering frames – – – – .636* .196 .623* .195
Antisocial behavior .123* .023 .126* .023 .120* .012 .121* .021
Success goals −.031 .156 −.020 .154 .073 .175 .055 .168
Family pregnancy reaction .125 .078 .125 .077 .122 .075 .118 .075
Pregnancy Impact .122 .080 .119 .077 .069 .079 .073 .078
Family structure .497* .218 .577* .210 .584* .194 .605* .190
Authoritative parenting −.282 .290 −.143 .292 −.030 .248 −.018 .246
Single .316 .219 .356 .216 .425* .206 .441* .208
Annual income .009 .008 .009 .008 .008 .007 .010 .007
Age (Months) .029* .013 .034* .012 .031* .012 .032* .012
Variance component .148 .146 .143 .143
*p < .05
Neighborhood Culture and Sexual Partnering 1837
exposure to the neighborhood player climate and the salience of player frames
increase the log-odds of having a higher number of sex partners.
To disentangle these patterns, we conduct a MOLM mediation analysis that
decomposes the direct and indirect effects of the neighborhood cultural climate.
The results of the mediation analysis are displayed in table 3. According to the
estimates, the effects of neighborhood partnering climate on sexual behavior are
partially mediated or indirect through individual frames (b = .465). In fact, indi-
vidual player frames mediate 28.5 percent of the total effects (b = 1.63) of the
neighborhood climate on sexual partnering. Moreover, the direct effects of the
neighborhood cultural climate are also signicant (b = 1.16) at the 95 percent
level (not shown in tabular form), indicating that exposure to neighborhood
player culture increases multiple sexual partnering, independent of its effects
through individual frames. Altogether, the ndings from the mediation decompo-
sition analysis suggest that young men accumulate multiple sex partners in part
to comply with the rules of the local neighborhood culture but also because they
assign strong personal meaning to player frames. Combined, the evidence from
the multilevel models indicates that contextual, along with mediation-type, pro-
cesses constitute the linkages between neighborhood cultural mechanisms and
sexual partnering.3
Finally, we examine the status-deprivation hypothesis, which predicts that
player frames have stronger effects on sexual behavior among youth in disadvan-
taged neighborhoods, whereas the effects should be weaker—or less predictive of
actual sexual behaviors—in low-poverty environments. Model 4 of table 2 dis-
plays results from a MOLM model that species a cross-level interaction between
neighborhood disadvantage and individual player frames. As seen in model 4, the
interaction term is signicant and positive (b = .315, SE = .149), which is sup-
portive of the deprivation hypothesis, suggesting that the effects of player frames
are more pronounced in disadvantaged neighborhoods. Figure2 displays the
multiplicative effects implied in the interaction term.4 Among young men who
strongly align with player frames, the log-odds of multiple sexual partnering are
signicantly higher if they reside in relatively disadvantaged neighborhoods (+1
SD above the mean). By contrast, among those who reside in more advantaged
neighborhoods (–1 SD below the mean) the log-odds of multiple partnering are
signicantly lower; moreover, the slope effect of frames is not nearly as steep.
Taken together, these results support the prediction that player frames are more
consequential for involvement in multiple-partnering outcomes for young men
residing in structurally disadvantaged neighborhoods.
Table3. MOLM Mediation Decomposition Results: Total and Indirect Effects of Neighorhood
Player Culture on Log-Odds of Multiple Sexual Partnering
Estimate 95% C.I.
Portion of Total that
is Indirect
Indirect effects via individual frames .465* (.141, .782) 28.51%
Total effects 1.63* (.492, 2.71)
*p < .05
1838 Social Forces 94(4)
Discussion and Conclusion
That neighborhood context is a signicant backdrop for adolescent sexual
experiences is increasingly clear; and the question of how neighborhoods inu-
ence dimensions of adolescent sexual conduct is beginning to be more fully
understood theoretically and empirically. Anderson’s (1999) inuential perspec-
tive, in concert with other leading neighborhood-effects models (e.g., Wilson
1996), provides a conceptual platform for expectations about the means
through which a player culture operates to affect sexual partnering. Building on
prior research (e.g., Giordano etal. 2009) and recent research in the sociology
of culture, this study advanced and tested a set of hypotheses pertaining to the
cultural context of sexual decision-making among young African American
males: We examined whether neighborhood player culture increases multiple
sexual partnering as a result of either a contextual or an indirect mechanism or
both. We then tested a status-deprivation hypothesis to determine whether
neighborhood deprivation has a multiplicative effect. This is among the rst
studies of multiple sexual partnering to examine contextual cultural processes
using theoretically consistent survey measures aggregated to the neighborhood
level.
The results from this study contribute to the literature on neighborhood effects
and the social context of adolescent sexual activities; they also highlight direc-
tions for future research. First, consistent with prior research, the study suggests
that neighborhood culture is a signicant source of variation in multiple sexual
partnering among African American adolescent males (see Warner, Giordano,
Manning and Longmore 2011). The results also yield key theoretical insights into
the processes through which the player culture affects sexual behavior. Speci-
cally, a contextual effect was observed in the analysis: Neighborhood player
Figure2. Effects of individual player frames on multiple sexual partnering across levels of
neighborhood structural deprivation
–1
–0.5
0
0.5
1
1.5
2
2.5
3
3.5
1 (low)2
34
(high)
Log odds of sexual behavior
Individual Partnering Frames
Low neighorhood deprivation (–1 SD)
Meadium neighorhood deprivation (Mean)
High neighborhood deprivation (+1 SD)
Neighborhood Culture and Sexual Partnering 1839
culture predicted multiple sexual partnering, controlling for individual-level
frames. Hence, this pattern suggests that young men have sex with multiple part-
ners to conform to messages of the neighborhood culture. For some men, there
may be status penalties for not acting in accordance with the player culture, and
therefore they do so mainly out of reputational concerns. Evidence of this contex-
tual effect is consistent with Anderson’s claim that young men pursue multiple
sex partners to satisfy local status imperatives. Additionally, however, individual-
level player frames predicted sexual partnering independent of the neighborhood
culture, which indicates that an indirect mechanism of frame resonance also is
involved. Estimates from the mediation analysis claried how the indirect process
appears to unfold: For some adolescents, exposure to the neighborhood player
culture increases the personal salience of partnering frames, which thereby
increases their odds of having numerous sex partners.
Altogether, neighborhood culture appears not only to entail the commonly
referenced indirect mechanism through frame salience, but it also places pressure
on individuals in the group to conform. It is noteworthy that similar conclusions
have been drawn in multilevel research on the cultural context of youth aggres-
sion (Bernburg and Thorlindsson 2005). More broadly, the ndings illuminate
the fact that adolescent sexual-partnering strategies are not merely a product of
individual differences in known risk factors; rather, emergent social processes
also affect how young men approach sexual partnering. Consistent with general
theoretical notions, the local neighborhood appears to constrain or enable sexual
decision-making above and beyond individual preferences.
Second, we also uncovered empirical support for the status-deprivation
hypothesis. Indeed, the player frame is more strongly related to the sexual-
partnering behaviors of males residing in more disadvantaged neighborhoods
than for those in low-poverty neighborhoods. From a broader perspective, these
patterns afrm Brewster, Billy, and Grady’s (1993, 716) general assertion that the
neighborhood “cultural environment and local opportunity structure are inter-
twined in complex ways.” These ndings also comport with ethnographic
research suggesting the economic opportunities to which youth are exposed
affect the social signicance of the player role (e.g., Harding 2010). It is not nec-
essarily the case that poor youth are more interested in sexual relations than
middle-class youth, or that the “sexual codes of inner-city youths … differ funda-
mentally from those expressed by young people in other settings” (Anderson
1989, 76). Rather, it is likely the case that the social and personal consequences
of adolescent sexual conduct vary across socioeconomic positions, which is partly
responsible for the stratication of sexual experiences. Among youth in impover-
ished neighborhoods, the pursuit of multiple sex partners may bring about more
prized social status; and at the same time, perhaps the risk of fathering a child
comes at less of a cost to their perceived prospects for economic attainment since
these opportunities are often limited in poor environments (see Harding 2010;
Wilson 1996). Elsewhere, research on fertility and contraceptive use also points
to the importance of examining how neighborhood poverty affects the nature of
adolescent romantic relationships including fertility (Edin and Kefalas 2005).
Perhaps the intersection of cultural mechanisms and structural deprivation
should be further integrated into theoretical discussions of sexual activity and
1840 Social Forces 94(4)
thegenerational persistence of inequality among low-income youth (see also
Furstenberg 2001).
Looking beyond the current study, there is a revival of interest in the implica-
tions of neighborhood culture for youth well-being and the reproduction of social
inequalities (Lamont and Small 2008). More recent conceptualizations of culture
as frames or scripts clarify the ways in which actors use culture selectively, as they
carry a diverse tool kit of models that guide their decision-making (see Berg etal.
2013). The conceptual specication of player frames employed in the current
study assumes that the signicance of cultural orientations is exible—they can
change—across the developmental life course. Indeed, as young men age, they
may come to disavow the status granted to player-like behavior—perhaps the
frame eventually loses its personal salience (for similar research, see Giordano
etal. 2009). Similarly, Harding’s (2007) framework of cultural heterogeneity
based on cognitive interpretations also claries the diversity of pregnancy frames
in poor communities, highlighting that conventional models are the dominant
orientation even in impoverished places. While our study was an explicit test of
the conceptual model implied in Anderson’s (1999) research and similar works,
it is conceptually attentive to script or cognitive perspectives of cultural processes
(e.g., Berg, Sevell, and Stewart 2015). Perhaps future research should examine the
degree to which the diversity of cultural models pertaining to player styles of
conduct affects multiple partnering and how this process inuences other com-
ponents of sexual decision-making (e.g., condom use).
The study is not without limitations. Although the samples are selected from
two locations in separate regions of the country, the ndings may not generalize
to adolescent males elsewhere in the country. Also, our measure of player frames
does not necessarily capture elements of adolescents’ personal identities per se,
given that the items inquired about general status standards; still, the strong rela-
tionship between frames and sexual behavior is indicative of the personal salience
of the frame as a resource for action. Another limitation arises from the fact that
we are unable to unequivocally establish the temporal association between sex-
ual-conduct norms and multiple partnering. It is possible that player frames were
galvanized after some youths initiated their sexual behaviors, meaning their
beliefs may be a product of their experiences rather than the reverse. However,
this possibility would not necessarily taint the contextual effect of partnering
climate because the contextual measure is statistically purged of a respondent’s
personal frames, meaning the contextual effect is signicantly less vulnerable to
reverse temporal effects in this case. Still, wave 3 measures of player frames are
not available in FACHS; however, to gain a better understanding of this potential
ordering issue, we conducted additional analysis using two analytical approaches.
We rst conducted a logistic regression analysis on an outcome assessing wave-
3-to-4 change in the ordinal rank of sex partners and, further, we incorporated a
proxy measure of pregnancy frames from wave 3 in a separate full analysis of the
original model specication; this proxy measure is meant to capture prior senti-
ments potentially aligned with the partnering culture.5 Each strategy brought
methodological advantages and limitations for assessing this temporal pattern-
ing. Altogether, the results from these additional analyses are parallel to those
reported in the original specication, providing further supplemental evidence
Neighborhood Culture and Sexual Partnering 1841
supporting the main ndings and study conclusions. Also, it is worth noting the
possibility that youth who espouse the player culture will falsely report higher
numbers of sexual experiences to project a favorable image. Although the current
data set is limited in these respects, it is worth noting that the data are uniquely
equipped to allow for research insights into neighborhood cultural and organiza-
tional processes thought to affect sexual development.
Taken together, this study contributes to the literature on adolescent sexual
behavior by examining how the cultural backdrop of neighborhoods guides
young males’ sexual-partnering practices; the results provide additional evidence
needed to evaluate hypotheses regarding the role of cultural processes. After a
period in which culture was marginalized in the research on adolescent develop-
ment, sociology has experienced a revival of interest in the explanatory role of
culture. More broadly, addressing the question of how neighborhood social pro-
cesses operate is important to resolving both practical and theoretical questions
about adolescent sexual development. Perhaps the ndings can also inform policy
efforts designed to curb potentially risky behaviors occurring in nonromantic
relationships, which can have long-term consequences for economic attainment
and well-being (Furstenberg 2001). For instance, perhaps programs that advise
youth about the health risks of multiple sexual partnering could potentially curb
this behavior but also temper its normative salience.
Notes
1. Sociological research on the neighborhood context of non-marital fertility also uses
aggregate socioeconomic measures as proxy measures of culture or control processes
(e.g., Sucoff and Upchurch 1998). Since the etiology of these outcomes may differ
from multiple sexual partnering, we are cautious about using this body of ndings to
derive assumptions about sexual partnering.
2. As Clarke (2008, 752) notes, “there are various rules of thumb often in the range of
15 to 30 per group … about the lower threshold at which data sparseness renders
multilevel models unreliable or even unnecessary.” He notes, however, that these rules
are not based on strong empirical evidence. Additional simulation research identies
cell-size thresholds at which two-level estimates are reliable and valid (see Clarke and
Wheaton 2007); by the standards of those analyses, the data structure for this study
is unlikely to produce biased estimates of between-unit variance from a random-
intercept model.
3. To assess another potential multiplicative process, we estimated an interaction term
between neighborhood culture and individual frames. The term was not statistically
signicant at conventional levels (b = .601, SE = .491). We also conducted additional
analysis where the individual-level partnering frames measure was specied as the out-
come variable; the results showed a signicant positive association between neighbor-
hood partnering cultural climate and partnering frames (Coef. = 1.10; S.E. = .348). Al so ,
the data are technically censored given that approximately 14 percent of the respondents
reported having zero sexual partners; but the ordered logit estimates do not account for
censored distributions. To determine the degree to which this censoring affects the esti-
mates, we re-estimated the models using a multilevel Tobit specication. Altogether, the
results from the Tobit model were similar to those reported in the ordered logit analysis,
including the mediation and moderating effects central to the hypotheses. Finally, we
examined the effects of a sample-based or compositional measure of neighborhood
1842 Social Forces 94(4)
deprivation as a substitute for the current Census-based measure. As with the partnering
climate variable, we built the compositional measure using only sample information.
Specically, we used data from each respondent pertaining to ve household socioeco-
nomic characteristics that approximate the Census socioeconomic indicators. Average
values of these indicators were aggregated to the neighborhood level; the indicators all
loaded on a single deprivation factor construct (Eigenvalue >2.5). According to the
regression models, the sample-based deprivation variable has non-signicant effects on
sexual partnering; furthermore, the association between the cultural climate measure
and sexual partnering remains statistically signicant net of this substitute measure.
4. A small number of units were in the “zero” category; hence, the x-axis displays the
next value of 1.
5. Since the sexual-partners variables at waves 3 and 4 are measured on ordinal scales,
this limited our ability to assess change in the number of partners among those in the
highest category: 7 or more partners. Further, although “pregnancy norms” may rep-
resent sound proxies for sexual partnering frames, the two constructs are conceptu-
ally distinct. Indeed, the perceived salience of pregnancy frames could be affected by
the nancial costs of multiple-partner childbearing.
About the Authors
Mark T. Berg is an Assistant Professor in the Department of Sociology at the
University of Iowa. His research interests include neighborhood cultural and
social organizational processes, the social psychology of interpersonal violence,
antisocial behavior and adolescent health and well-being.
Callie H. Burt is an Assistant Professor in the Department of Sociology at the
University of Washington. Her primary research interests are in developmental
and life-course criminology and social stratication. Her work focuses on eluci-
dating the social, psychological, and, recently, biological mechanisms through
which structural and cultural stressors and supports inuence risky and/or anti-
social behaviors across the life course. Recent publications on these topics have
appeared in Criminology.
Man-Kit Lei is an assistant research scientist at the Center for Family Research
and an adjunct assistant professor in the Sociology Department at the University
of Georgia. His current research focuses on epigenetics, neighborhood processes,
and individual well-being. His work has appeared in the Journal of Family Psy-
chology, Social Science & Medicine, and American Sociological Review.
Leslie Gordon Simons is a Professor of Sociology at the University of Georgia.
Her program of research focuses on the socio-contextual predictors and conse-
quences of various family processes as well as the mediators and moderators the
relationship between experiences in the family of origin and outcomes for adoles-
cents and emerging adults. Specically, she examines the intergenerational trans-
mission of problem behaviors and the mechanisms that link parenting to
behavioral outcomes such as delinquency, intimate-partner violence, and risky sex.
Eric A. Stewart is a Professor in the College of Criminology and Criminal Jus-
tice at Florida State University. He is a member of the Racial Democracy, Crime,
and Justice Network. His research interests include racial inequality and criminal
outcomes, crime over the life course, and contextual processes and micropro-
cesses that affect adolescent development.
Neighborhood Culture and Sexual Partnering 1843
Ronald Simons is a Distinguished Research Professor in the Department of
Sociology at the University of Georgia. His research has examined the develop-
mental roots of problem behavior as well as the extent to which adult transitions
involving romantic relationships, employment, and incarceration inuence desis-
tance. Recently, his research program has expanded to include tests of various
biosocial models.
References
Anderson, Elijah. 1989. “Sex Codes and Family Life among Poor Inner-City Youths.” Annals of the American
Academy of Political and Social Science 50(1):59–78.
______. 1999. Code of the Street: Decency, Violence, and the Moral Life of the Inner City. NewYork: W. W.
Norton Press.
Baumer, Eric, and Scott J. South. 2001. “Community Effects on Youth Sexual Activity.” Journal of Marriage
and Family 63(2):540–54.
Benford, Robert, and David Snow. 2000. “Framing Processes and Social Movements: An Overview and
Assessment.” Annual Review of Sociology 26(1):611–39.
Berg, Mark T., Eric A. Stewart, Endya Stewart, and Ronald L. Simons. 2013. “A Multilevel Examination of
Neighborhood Social Processes and College Enrollment.” Social Problems 60(4):513–34.
Berg, Mark T., Eric A. Sevell, and Eric A. Stewart. 2015. “Cultural Processes, Social Order, and Criminology.”
In, The Handbook of Criminological Theory, edited by Alex R. Piquero, UK: Wiley-Blackwell.
Bernburg, Jon Gunnar, and Thorolfur Thorlindsson. 2005. “Violent Values, Conduct Norms, and Youth
Aggression: A Multilevel Study in Iceland.” Sociological Quarterly 46(3):457–78.
Brewster, Karin L., John O. G. Billy, and William R. Grady. 1993. “Social Context and Adolescent Behavior:
The Impact of Community on the Transition to Sexual Activity.” Social Forces 71(3):713–40.
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(5):758–78.
Browning, Christopher R., and Matisa Olinger-Wilbon. 2003. “Neighborhood Structure, Social Organization,
and Number of Short-Term Sexual Partnerships.” Journal of Marriage and Family 65(3):730–45.
Carlson, Marcia J., and Frank F. Furstenberg. 2006. “The Prevalence and Correlates of Multipartnered
Fertility among Urban US Parents.” Journal of Marriage and Family 68(3):718–32.
Cialdini, Robert, and Noah J. Goldstein. 2004. “Social Influence: Compliance and Conformity.” Annual
Review of Psychology 55:591–62.
Clarke, Philippa. 2008. “When Can Group-Level Clustering Be Ignored? Multilevel Models versus Single-
Level Models with Sparse Data.” Journal of Epidemiology and Community Health 62(8):752–58.
Clarke, Philippa, and Blair Wheaton. 2007. “Addressing Data Sparseness in Contextual Population Research
Using Cluster Analysis to Create Synthetic Neighborhoods.” Sociological Methods & Research
35(3):311–51.
Crane, Jonathan. 1991. “The Epidemic Theory of Ghettos and Neighborhood Effects on Dropping Out and
Teenage Pregnancy.” American Journal of Sociology 96:1226–59.
Cubbins, Lisa A., and Koray Tanfer. 2000. “The Influence of Gender on Sex: A Study of Men’s and Women’s
Self-Reported High-Risk Sex Behavior.” Archives of Sexual Behavior 29(3):229–57.
DiMaggio, Paul. 1997. “Culture and Cognition.” Annual Review of Sociology 23:263–87.
Dupere, Veronique, Eric Lacourse, J. Douglas Willms, Tama Leventhal, and Richard E. Tremblay. 2008.
“Neighborhood Poverty and Early Transition to Sexual Activity in Young Adolescents: A Developmental
Ecological Approach.” Child Development 79(5):1463–76.
Edin, Kathryn, and Maria Kefalas. 2005. Promises I Can Keep: Why Poor Women Put Motherhood Ahead of
Marriage. Berkeley: University of California Press.
1844 Social Forces 94(4)
Furstenberg, Frank F. Jr., S. Philip Morgan, Kristin A. Moore, and James L. Peterson. 1987. “Race Differences
in the Timing of Adolescent Intercourse.” American Sociological Review 52(4):511–18.
Furstenberg, Franklin. 2001. “The Fading Dream: Prospects for Marriage in the Inner City.” In Problem of the
Century: Racial Stratification in the United States, edited by Anderson, Elijah, and Douglas Massey.
NewYork: Russell Sage Foundation.
Gibbons, Frederick X., Meg Gerrard, and David J. Lane. 2003. “A Social-Reaction Model of Adolescent
Health Risk.” In Social Psychological Foundations of Health and Illness, edited by Sulls, Jerry and
Kenneth Wallston. New York: Blackwell Publishing.
Giordano, Peggy. 2003. “Relationships in Adolescence.” Annual Review of Sociology 29:257–81.
Giordano, Peggy C., Monica A. Longmore, Wendy D. Manning, and Miriam J. Northcutt. 2009. “Adolescent
Identities and Sexual Behavior: An Examination of Anderson’s Player Hypothesis.” Social Forces
87(4):1813–1843.
Goffman, Erving. 1974. Frame Analysis: An Essay on the Organization of Experience. Cambridge, MA:
Harvard University Press.
Green, Traci C., Enrique R. Pouget, Magdalena Harrington, Faye S. Taxman, Anne G. Rhodes, Daniel
O’Connell, Steven S. Martin, Michael Prendergast, and Peter D. Friedmann. 2012. “Limiting Options: Sex
Ratios, Incarceration Rates, and Sexual Risk Behavior among People on Probation and Parole.” Sexually
Transmitted Diseases 39(6):424–37.
Halpern, Carolyn, Kara Tucker, J. Joyner, Richard Udry, and Chirayath Suchindran. 2000. “Smart Teens Don’t
Have Sex (or Kiss Much Either).” Journal of Adolescent Health 26(3):213–25.
Hannerz, Ulf. 1969. Soulside: Inquiries into the Ghetto Culture and Community. New York: Columbia Univer-
sity Press.
______. 2007. “Cultural Context, Sexual Behavior, and Romantic Relationships in Disadvantaged Neighbor-
hoods.” American Sociological Review 72(2):341–64.
Harding, David J. 2010. Living the Drama: Community, Conflict, and Culture among Inner-City Boys. Chicago:
University of Chicago Press.
Hays, Sharon. 1994. “Structure and Agency and the Sticky Problem of Culture.” Sociological Theory
12(3):57–72.
Jencks, Christopher, and Sharon Mayer. 1990. “The Social Consequences of Growing Up in a Poor Neighbor-
hood.” In Inner-City Poverty in the United States, edited by Lynn, L. E. Jr., and M. G. H. McGeary, 111–86.
Washington, DC: National Academies Press.
Krull, Jennifer L., and David P. MacKinnon. 2001. “Multilevel modeling of individual and group level medi-
ated effects.” Multivariate Behavioral Research 36(2):249–77.
Kornhauser, Ruth. 1978. Social Sources of Delinquency. Chicago: University of Chicago Press.
Lamont, Marc, and Mario Small. 2008. “How Culture Matters: Enriching Our Understanding of Poverty.” In
The Colors of Poverty: Why Racial and Ethnic Disparities Persist, edited by Harris, David, and Ann Lin,
76–102. New York: Russell Sage.
Lauritsen, Janet. 1994. “Explaining Race and Gender Differences in Adolescent Sexual Behavior.” Social
Forces 72(3):859–83.
Leventhal, Tama, and Jeanne Brooks-Gunn. 2000. “The Neighborhoods They Live In: The Effects of Neigh-
borhood Residence on Child and Adolescent Outcomes.” Psychological Bulletin 126(2):309–29.
Lopez, Vera. 2014. “Love Is a Battlefield: Mexican American Girls’ Strategies for Avoiding Players.” Youth
and Society 30(1):1–23.
Manning, Wendy D., Monica A. Longmore, and Peggy C. Giordano. 2005. “Adolescents’ Involvement in
Non-Romantic Sexual Activity.” Social Science Research 34(2):384–407.
Massey, Douglas, and Nancy Denton. 1993. American Apartheid: Segregation and the Making of the Under-
class. Cambridge, MA: Harvard University Press.
McCarthy, Bill, and Eric Grodsky. 2011. “Sex and School: Adolescent Sexual Intercourse and Education.”
Social Problems 58(2):213–34.
Neighborhood Culture and Sexual Partnering 1845
Mollborn, Stephanie. 2010. “Predictors and Consequences of Adolescents’ Norms against Teenage Preg-
nancy.” Sociological Quarterly 51(2):303–28.
Paik, Anthony, and Vernon Woodley. 2012. “Symbols and Investments as Signals: Courtship Behaviors in
Adolescent Sexual Relationships.” Rationality and Society 24(1):3–36.
Preacher, Kristopher J., Michael J. Zyphur, and Zhen Zhang. 2010. “A General Multilevel SEM Framework
for Assessing Multilevel Mediation.” Psychological Methods 15(3):209–21.
Roche, Kathleen M., Debra Mekos, Cheryl S. Alexander, Nan Marie, Karen Astone, Bandeen-RocheEns-
mingerMargaret E.. 2005. “Parenting Influences on Early Sex Initiation among Adolescents: How Neigh-
borhood Matters.” Journal of Family Issues 26(1):32–54.
Sampson, Robert J. 2012. Great American City: Chicago and the Enduring Neighborhood Effect. Chicago:
University of Chicago Press.
Sampson, Robert J., Stephen W. Raudenbush, and Felton Earls. 1997. “Neighborhoods and Violent Crime:
A Multilevel Study of Collective Efficacy.” Science 277(5328):918–24.
Sampson, Robert J., and William J. Wilson. 1995. “Toward a Theory of Race, Crime, and Urban Inequality.”
In Race, Crime, and Justice: A Reader, edited by Peterson, R., and J. Hagan, Pp. 37–54. New York: NYU
Press.
Santelli, John S., Nancy D. Brener, Richard Lowry, Amita Bhatt, and Laurie S. Zabin. 1998. “Multiple Sexual
Partners among US Adolescents and Young Adults.” Family Planning Perspectives 30(6):271–75.
Shaw, Clifford, and Henry McKay. 1969. Juvenile Delinquency and Urban Areas, rev. ed. Chicago: University
of Chicago Press.
Simons, Ronald L., Man Kit, Steven R. H. Lei, Gene H. Beach, Robert A. Brody, PhilibertGibbonsFrederick X..
2011. “Social Environmental Variation, Plasticity Genes, and Aggression: Evidence for the Differential
Susceptibility Hypothesis.” American Sociological Review 76(6):833–53.
Simons, Ronald, Leslie Gordon Simons, Callie H. Burt, Gene H. Brody, and Carolyn Cutrona. 2005. “Collective
Efficacy, Authoritative Parenting, and Delinquency: A Longitudinal Test of a Model Integrating Commu-
nity- and Family-Level Processes.” Criminology 43(4):989–1029.
Small, Mario L. 2004. Villa Victoria: The Transformation of Social Capital in a Boston Barrio. Chicago:
University of Chicago Press.
Small, Mario, David Harding, and Michele Lamont. 2010. “Reconsidering Culture and Poverty.” Annals of
the American Academy of Political and Social Sciences (May):6–28.
Small, Mario L., and Katherine Newman. 2001. “Urban Poverty after the Truly Disadvantaged: The Redis-
covery of the Family, Neighborhood, and Culture.” Annual Review of Sociology 27(1):23–45.
Sucoff, Clea A., and Dawn M. Upchurch. 1998. “Neighborhood Context and the Risk of Childbearing among
Metropolitan-Area Black Adolescents.” American Sociological Review 63(4):571–85.
Swidler, Anne. 1986. “Culture in Action: Symbols and Strategies.” American Sociological Review 51(1 ):27 3–86.
Vaisey, Stephen. 2008. “Socrates, Skinner, and Aristotle: Three Ways of Thinking about Culture in Action.”
Sociological Forum 23(3):603–13.
Warner, Tara D., Peggy C. Giordano, Wendy D. Manning, and Monica A. Longmore. 2011. “Everybody’s Doin’
It (Right?): Neighborhood Norms and Sexual Activity in Adolescence.” Social Science Research
40(6):1676–90.
Warner, Tara D., Wendy D. Manning, Peggy C. Giordano, and Monica A. Longmore. 2011. “Relationship
Formation and Stability in Emerging Adulthood: Do Sex Ratios Matter?” Social Forces 90(1):269–95.
Wilson, William J. 1996. When Work Disappears: The World of the New Urban Poor. New York: Vintage
Books.
1846 Social Forces 94(4)
