General Strain Theory, Gender, and the Conditioning Influence of Negative Internalizing Emotions on Youth Risk Behaviors

• DOI: 10.1177/1541204013477428
• Published date: 01 January 2014
• Date: 01 January 2014
• Subject Matter: Articles

Article
Youth Violence and Juvenile Justice
2014, Vol 12(1) 58-76
General Strain Theory,
ª The Author(s) 2013
Reprints and permission:
Gender, and the Conditioning
sagepub.com/journalsPermissions.nav
DOI: 10.1177/1541204013477428
yvj.sagepub.com
Influence of Negative
Internalizing Emotions on
Youth Risk Behaviors
Kimberly A. Francis1
Abstract
Girls’ maladaptive responses to strain may be more likely to manifest in self-directed deviance than
externally directed deviance, partly due to the role of depression/anxiety in girls’ lives. These asser-
tions are tested using the Project on Human Development in Chicago Neighborhoods. Gendered
maladaptive outcomes—aggressive delinquency, running away, substance use, and suicidal
behavior—are regressed on negative emotions and serious strain. The moderating effects of depres-
sion/anxiety are also tested. Depression/anxiety is associated with less aggressive delinquency for
girls and amplifies the effects of strain and anger on nonaggressive maladaptive outcomes. The
results help explain how gender influences relationships among strain, externalizing and internalizing
negative emotions, and risk behaviors.
Keywords
general strain theory, gender and deviance, negative emotions
General strain theory (GST) posits that exposure to certain stressors or strains increases the
likelihood for criminal behavior through increases in a range of negative emotions, especially anger;
individuals engage in criminal behavior as a way to cope and alleviate the negative emotions
(Agnew, 1992, 2006). Early gender-focused tests of GST established that the direct relationship
between strain and delinquency is positive and consistent across sex (Broidy, 2001; Hoffman &
Su, 1997; Mazerolle, 1998). A body of literature soon emerged that explored how GST might
explain gender differences in delinquency. The ability of GST to explain why girls are less likely
than boys to cope with strain with law-violating behaviors remains an open question, especially
since women and girls are exposed to more stressors than their male counterparts (Bush & Simmons,
1987; Daigle, Cullen, & Wright, 2007; Mirowsky & Ross, 1995; Turner, Wheaton, & Lloyd, 1995),
1 Abt Associates, Cambridge, MA, USA
Corresponding Author:
Kimberly A. Francis, Abt Associates, 55 Wheeler Street, Cambridge, MA 02138, USA.
Email: kimberly_francis@abtassoc.com

---

Francis
59
and women and girls tend to experience equal or greater levels of anger as well as greater levels of
co-occurring negative emotions with anger, like depression, guilt, anxiety, fear, and frustration
(Broidy, 2001; De Coster & Zito, 2010; Hay, 2003; Jang, 2007; Mirowsky & Ross, 1995; Piquero
& Sealock, 2004; Sharp, Brewster, & Redhawk-Love, 2005; Van Gundy, 2002).
Building on Broidy and Agnew (1997), recent tests of GST have examined the extent to which
girls are more likely to respond to strain with self-directed and escapist deviant behaviors, such as
disordered eating, substance use, suicide attempts, and running away, than externally focused
deviance due to particular combinations of strains and negative emotions that are more conducive
to these behaviors (Broidy, 2001; Jang, 2007; Kaufman, 2009; Piquero, Fox, Piquero, Capowich,
& Mazerolle, 2010; Sharp, Terling-Watt, Atkins, Gilliam, & Sanders, 2001, 2005). These studies
test the idea that boys are more likely than girls to experience anger alone in response to strain,
which increases the risk of externally focused delinquency; girls, on the other hand, are thought
to experience similar levels of anger to boys in conjunction with internalizing negative emotions
such as depression, anxiety, and guilt. The presence of these non-angry negative emotions are
thought to curb the influence of anger on externally focused delinquency but increase the risk of less
aggressive and more internally focused problem behaviors. Negative emotions are thus pivotal in
explaining gender differences in responses to strain, and more research is needed to disentangle the
gendered relationships between different types of strain, negative emotions, and coping responses
(Agnew, 2006).
In particular, the role of depression and other negative internalizing emotions in GST is unclear.
One study (Broidy, 2001) found that non-angry negative emotions in conjunction with feelings of
anger decrease the likelihood of delinquency for young women and not young men. Others have
found either a positive relationship between depression and delinquency for both young women and
men (Kaufman, 2009) or young men only (De Coster & Zito, 2010), or no relationship for either sex
(Piquero & Sealock, 2004). Negative emotions in prior studies are usually conceptualized as
mediating variables that account for the strain–delinquency relationship. Few studies have addressed
whether they might influence the effects of strain and anger through an interaction effect. That is, if
anger and negative internalizing emotions are co-occuring, how do they predict qualitatively differ-
ent problem behaviors? What if negative internalizing emotions interact with strain and anger to
either amplify or mitigate their effects on gendered outcomes?
This article tests an expanded model of GST that focuses on the role of internalizing negative
emotion (i.e., depression/anxiety) as a potential moderator of both strain and anger to explain
gendered responses to strain. Using data from a stratified random sample of adolescents in Chicago
households, the test includes measures of serious strain salient for both adolescent males and
females, measures of both anger and negative internalizing emotions, and an expanded scope of the
dependent variable to include aggressive delinquency, running away, substance use, and suicidal
behavior. It is one of the first papers to position internalizing negative emotion as a conditioning
factor that interacts with multiple strains and anger to influence the likelihood of different gendered
problem behaviors.
Gender and Negative Emotions in GST
Broidy and Agnew (1997) propose that girls and boys should be equally likely to respond to strain
with anger; but girls will be more likely to experience depression, which in turn reduces the likeli-
hood of aggressive delinquency and increases the likelihood of inner-directed problem behaviors.
Tests of Broidy and Agnew’s hypotheses generally suggest that the relationship between strain and
negative emotions varies depending on the type of outcome under consideration (Ganem, 2010;
Jang, 2007; Kaufman, 2009; Piquero & Sealock, 2004; Sharp et al., 2005), and girls may experience
an array of negative emotions that reduces the likelihood of delinquency but promotes other types of

---

60
Youth Violence and Juvenile Justice 12(1)
maladaptive coping (Jang, 2007; Sharp et al., 2001, 2005). The first set of hypotheses tests the extent
to which the effects of strain on anger and depression/anxiety varies by sex.
Hypothesis 1a: The relationship between strain and anger does not vary by sex.
Hypothesis 1b: The relationship between strain and depression/anxiety is stronger for girls than boys.
The second hypothesis builds on the first to test whether negative internalizing emotion has a
negative effect on aggressive delinquency and a positive effect on nonaggressive problematic coping
behaviors, net of anger, strain, and background factors. GST distinguishes between ‘‘state’’ negative
emotions, or those felt at a certain point in time in response to the strainful event, and ‘‘trait’’ neg-
ative emotions, or a predisposition to feel angry, depressed, anxious, or frustrated (Agnew, 2006;
Mazerolle, Piquero, & Capowich, 2003). GST expects that ‘‘state’’ emotions should account for the
strain-deviance relationship, while ‘‘trait’’ emotions are less likely to fully mediate this relationship
and may have an independent direct effect on deviance (Mazerolle et al., 2003). Moreover, ‘‘trait’’
emotions may be best conceptualized as conditioning factors in the GST process (Agnew, 2006;
Manasse & Ganem, 2009). Most GST research relies on measures of ‘‘trait’’ emotion as a proxy due
to the frequent inavailability of ‘‘state’’ emotion measures in extant data, a limitation of studies that
test the mediating role of negative emotions. For these reasons, trait emotion measures are not
expected to fully account for the strain–deviance relationship in the second hypothesis, and their role
as a conditioning variable is highlighted in the third hypothesis.
Agnew (2006) singles out anger as the negative emotion most likely to account for the effect of
strain and lead to delinquency outcomes because it motivates a person to take action to right a
perceived injustice, is associated with blaming others for negative experiences, and creates a desire
for revenge. Accordingly, anger is more likely to predict aggressive behaviors, such as fighting, than
nonaggressive behaviors like drug use and some property crime (Aseltine, Gore, &Gordon, 2000;
Brezina, 1996, 1998; Broidy, 2001; Mazerolle & Piquero, 1998).
There is a clear positive relationship between undesirable life events or traumas and depression
(Mirowsky & Ross, 2003; Turner & Lloyd, 1995). However, Agnew (1992, 2006) hypothesizes that
depression should be less strongly related to aggressive delinquency and more strongly related to
drug use and other nonaggressive, internally directed behaviors. Jang (2007) and Jang and Johnson
(2003) found support for these relationships in their study of a nationally representative sample of
African Americans, as did Sharp et al. (2005) in a localized study of...