Subcultural Values and Violent Delinquency

• Author: Pamela Wilcox,Graham C. Ousey
• Published date: 01 January 2005
• Date: 01 January 2005
• DOI: http://doi.org/10.1177/1541204004270942
• Subject Matter: Articles

10.1177/1541204004270942
Youth Violence and Juvenile Justice
Ousey, Wilcox / SUBCULTURE OF VIOLENCE TRADITION
SUBCULTURAL VALUES AND
VIOLENT DELINQUENCY
A Multilevel Analysis in Middle Schools
Graham C. Ousey
University of Delaware
Pamela Wilcox
University of Cincinnati
This study extends research in the subculture of violence tradition by examining the
simultaneous effects of individual- and aggregate-level proviolence values on violent
behavior. With data from a survey of 3,690 seventh-grade students nested within 65
Kentucky schools, we found that an initially observed between-school effect of violent
values on in-school violent behavior is not a contextual effect. Rather, the between-
school effect of violent values is primarily a reflection of the individual-level association
between violent values and violent behavior. Moreover, we found that the effect of indi-
vidual values remains evident after measures of impulsivity and exposure to violent
peers are controlled. No evidence of cross-level interaction effects involving individual
values and the school-level violent value context was found.
Keywords: subculture of violence; social learning; violence; delinquency; schools
One classic explanation of violent crime suggests that individual differences in vio-
lent offending are due, in part, to differences in the extent to which individuals are exposed
to and have internalized a value system that tolerates, condones, or even demands violent
responses to various social situations. In general, this subculture of violence perspective
posits that individuals embedded in social contexts that feature social groups espousing
proviolence values will have a heightened proclivity for adopting these values and behaving
accordingly. Generally consistent with this theoretical model, a number of studies report
that rates of violence are higher in geographic localities or social groups in which pro-
violence values or attitudes are believed to be particularly prominent (e.g., see Curtis, 1975;
Gastil, 1971; Hackney, 1969; Messner, 1983; Wolfgang & Ferracuti, 1967). Moreover,
other studies report that individuals with more proviolence attitudes engage in violent
Authors’ Note: An earlier version of this article was presented at the 2002 meeting of the American Society of
Criminology, Chicago. The research was supported in part by grants from the National Institute on Drug Abuse
(DA-05312, DA-11317). We would like to thank Richard R. Clayton, Good Samaritan Foundation chair of Public
Health at University of Kentucky, and Bill Scott, the Kentucky School Boards Association, for their support
regarding this project.
Youth Violence and Juvenile Justice, Vol. 3 No. 1, January 2005 3-22
DOI: 10.1177/1541204004270942
© 2005 Sage Publications
3

---

4 Youth Violence and Juvenile Justice
behavior at a higher rate or frequency than those with fewer proviolence attitudes (Felson,
Liska, South, & McNulty, 1994; Heimer, 1997; Markowitz & Felson, 1998).
Despite this supporting evidence, the body of research in the subculture of violence
tradition contains a number of limitations. First, many of the studies are macro-level de-
signs that contain no explicit measures of violent subcultures (e.g., Gastil, 1971; Hackney,
1969; Messner, 1983; but see Warner, 2003). Rather, these studies simply infer subcultural
effects based on observing a statistically significant relationship between rates of violence
and region-indicator (e.g., South) or racial-composition (e.g., percentage Black) variables.
Second, extant studies that do incorporate explicit measures of the subculture of violence,
such as attitudes or values regarding the acceptability of violent behavior, have been framed
commonly at the individual level of analysis only (Heimer, 1997; Markowitz & Felson,
1998). Consequently, they overlook the fact that subcultures are emergent properties of
social groups that conceivably can affect violence directly (i.e., via a contextual main
effect), indirectly (e.g., by increasing the likelihood of acquiring violent attitudes and/
or violent peer associates), or by moderating the relationship between violence and its
individual-level correlates (i.e., cross-level interaction). Finally, despite making a seminal
contribution, the only multilevel subculture of violence study in the literature (Felson et al.,
1994) does not control for two theoretically important and empirically robust variables—
violent peer exposure and impulsivity or low self-control—when considering the relation-
ship between subcultural values and violence. Consequently, the degree to which group-
and individual-level variations in violent values affect violent behavior may be overstated.
In the current study, we build on the extant research literature on the subculture of
violence in a number of ways. Conceptualizing schools as social contexts that vary in the
degree to which violent subcultures exist, we utilized data from a sample of 3,690 seventh-
grade students nested within 65 Kentucky middle schools to estimate multilevel models of
the violent subculture–violent behavior nexus. Moreover, we simultaneously estimated the
effects on individual school-based violent behavior of aggregate- (i.e., school) and person-
level indicators of violent subcultural values. Next we examined whether the multilevel
effects of proviolence values remain robust in models that control for the influence of
important covariates (exposure to violent peers, impulsivity or low self-control) that are
generally omitted from prior studies. Finally, we examined the extent to which the mag-
nitude of the individual-level relationship between violent behavior and violent values
depends on the degree to which emergent, school-level violent subcultures are evident.
Theoretical Background
Subculture of Violence
The presumed link between violent behavior and proviolence values or attitudes has
roots in a number of theoretical traditions. Most notably, the subculture of violence theory,
exemplified by the work of Wolfgang and Ferracuti (1967), suggests that race and class dif-
ferences in rates of violence are a function of differences in the extent to which these groups
adhere to value systems that support or excuse the use of violence in various social situa-
tions. Following on this model, a number of studies have tested hypotheses related to the
subculture of violence at either the macro or micro level of analysis.
Macro-level work in this tradition has gone in two directions. The first, which can be
traced to the work of H. V. Redfield (1880), examined whether the comparatively high rates

---

Ousey, Wilcox / SUBCULTURE OF VIOLENCE TRADITION 5
of homicide and assault in the southern United States are the result of unique cultural values
that characterize that particular region of the country. Indeed, some argued that the histori-
cally high rates of lethal violence in the South are a product of a siege mentality that devel-
oped in response to non-Southerner’s frequent criticism of the region’s peculiarities, such
as its legacy of slavery and continued dependence on agriculture (Hackney, 1969). Still oth-
ers pointed to a Southern so-called frontier mentality in which grievances are expected to be
handled informally via duels, brawls, and feuds (Brearly, 1932; Bruce, 1979; J. S. Reed,
1982). Whatever the origin of this value system, the basic implication is that in the South
violence is condoned (or at least tolerated) in a wider range of situations than in regions
where this subculture of violence is not evident (Ousey, 2000). Consequently, rather than
turning the other cheek, Southerners may be more likely to respond to perceived insults or
injustices in ways that will escalate conflicts into potentially serious violent encounters.
The second line of macro-level research develops a similar argument with respect to
race. Drawing heavily on Wolfgang and Ferracuti’s (1967) original work, several studies
(e.g., Sampson, 1985) examine the hypothesis that the often-reported positive relationship
between rates of criminal violence and the percentage of the population that is Black is a
reflection of the existence of a race-based subculture of violence. Presumably this subcul-
ture heightens violent proclivities among African Americans by defining violence as appro-
priate or necessary in a wider range of social circumstances than is the case for other racial
groups (e.g., see Curtis, 1975).
In line with the basic expectations of the region- and race-based subculture of vio-
lence arguments, a number of macro-level studies suggests that rates of violence (partic-
ularly homicide) are higher in the South and/or among African Americans (Blau & Blau,
1982; Curtis, 1975; Gastil, 1971; Hackney, 1969; Messner, 1983). In contrast, other macro-
level studies indicate that these regional and racial differences in violence rates are ex-
plained by disparities in economic disadvantage (e.g., see Bailey, 1984; Corzine & Huff-
Corzine, 1992; Harer & Steffensmeier, 1992; Williams, 1984). In other words, higher rates
of violence among Southerners and Blacks emanate from the disadvantaged structural con-
texts endured by these groups, rather than from presumed differences in values or attitudes
regarding violence.
Although valuable as a starting point, the overall contribution of studies in this
macro-level genre is limited by (at least) two factors. First, these studies tend to develop a
rather shortsighted view of the relationship...