Читать книгу The Invisible Woman - Joanne Belknap - Страница 35

Agnew (1985, 1992) revised TST into general strain theory (GST). GST advances and expands earlier strain theories by broadening the sources and types of adaptations to strains and acknowledging that goals may vary depending on an individual’s gender, race, and class. Rather than simply focusing on structural factors limiting financial success (like TST), GST includes three psychosocial strain sources: (1) the presence of negative stimuli, (2) the loss of positive stimuli, and (3) the failure to achieve positive goals. According to GST, whether responses to strain and frustrations are law-abiding or delinquent depends on an individual youth’s personality, self-esteem, social support system, and so on (e.g., if anger is the response, the coping strategy is more likely to be delinquent) (Agnew 1992). Additionally, GST addresses the importance of allowing for varied goals due to individuals’ gender, race, and class differences (Broidy, 2001). Stated another way, GST suggests that both strains and the responses to these strains may be gendered, raced, or classed. Broidy and Agnew (1997) purported the gender gap in offending could also be due to gender differences in the types of strains and gender differences in the emotional responses to strains.

Bottcher (2001, p. 894) criticized GST for failing to consider gender as “a product of individual and interpersonal action,” and Agnew (2001) himself published concerns with the tests of GST, specifically that many key strains outlined in GST were not included in the tests and that most GST tests focused on a single, cumulative measure of strain. For example, he noted that child abuse (including sexual abuse) and criminal victimization are important to account for as stressors for delinquent behavior. Additionally, Agnew pointed out that it is necessary to look at additional characteristics of the strain: The more severe, unjust, lasting, and central to the individual’s life the strain is, the more likely it will result in anger, and thus, criminal behavior. Moreover, he recognized that abuse and criminal victimization are often perceived as unjust and serious and thus could result in stronger feelings of anger and injustice than other strains. Stated alternatively, victims of abuse may engage in delinquent or criminal behavior in efforts to compensate for the serious injustices they have experienced (Agnew, 2001, 2002).

Significantly, a 2008 review of GST (Cernkovich, Lanctôt, & Giordano, 2008) criticizes the tests of it for using almost exclusively male samples. Many recent GST studies have boy-only samples (e.g., Del Toro et al., 2019). Even when girls, women, or both are included, gender is used as a control variable rather than to understand how strains are gendered (e.g., Capowich, Mazerolle, & Piquero, 2001; C. Farrell & Zimmerman, 2018; Gallupe & Baron, 2009; Hay, 2003; Hay & Evans, 2006; Jang & Rhodes, 2012; M. C. Johnson & Menard, 2012; Langton & Piquero, 2007). In addition, abuse and trauma victimizations are far too often left out of the “strain” measures or analyses (e.g., Capowich et al., 2001; Cheung & Cheung, 2010; Dierenfeldt, Shadwick, & Kwak, 2019; D. Eitle, 2002; Hoffmann & Su, 1997). A GST study focusing on “family strain” did not include sexual abuse victimization (Hay, 2003).

Broidy’s (2001) test of GST reported that while strain causes anger in both sexes, girls were more likely to report other negative emotions (e.g., guilt, worthlessness, disappointment, depression, worry, fear, and insecurity). Broidy and Agnew (1997) found that both strain and the responses to strain explain gender differences in offending. For example, compared with boys, girls reported more restrictions on their lives and behaviors and greater family caretaking expectations; they were also more likely to report all types of abuse (emotional, physical, and sexual). While girls reported feeling more stress surrounding close relationships with friends and family, boys reported feeling more strain about external achievement such as material success. Another study found, as predicted by GST, that anger was a significant predictor of violent, property, and drug crimes, and criminal behavior was related to sexual abuse, homelessness, relative deprivation, and more deviant peers (Baron, 2004). Although this study reported that gender “was a significant predictor of crime” (Baron, 2004, p. 474), it did not explain how.

Notably, tests of GST often find many gender similarities in responses to the same strains. One study confirmed GST, finding that stressful life events increased the likelihood of delinquency, but this relationship was the same regardless of a youth’s gender, class, self-esteem, or perceived control over her or his environment (Hoffmann & Cerbone, 1999). Similarly, another GST study found that individual strains (i.e., physical abuse, sexual abuse, academic problems, future expectations, school dropout, and criminal legal system involvement) increased all three types of offending studied (property crimes, threatening interpersonal aggression, and using interpersonal aggression), and the increase was consistent across gender (Jennings, Piquero, Gover, & Perez, 2009).

A large GST study of youths who were referred to juvenile court looked at how gender and living situation were related to initial and recidivate arrests and youths’ responses to strain regarding drug offending (Grothoff, Kempf-Leonard, & Mullins, 2014). Girls reported 3 times as much child abuse as boys (physical, sexual, and emotional), and while it increased girls’ drug arrests (as expected), it decreased boys’. Boys, but not girls, not living with one or both parents were more likely to recidivate. Mental health problems increased both girls’ and boys’ drug offenses similarly (Grothoff et al., 2014). Hay’s (2003) GST study measured family strain using five dimensions: physical punishment, parental rejection, psychological control, unfair discipline, and non-intact family among high school students. Hay found no gender differences in youth anger levels resulting from family strain; however, girls’ feelings of guilt associated with family strain were higher than boys’. Moreover, this gender difference in the response of guilt to family strain explained much of the overall gender difference in delinquency, in that anger encourages delinquency while guilt discourages delinquency.

Notably, the GST tests have focused far more on anger than depression. Many studies conclude that boys report significantly more delinquent behavior than girls do, whereas girls report more negative self-feelings (e.g., depression, anxiety, self-esteem) than boys do (Jang & Rhodes, 2012; Jennings et al., 2009; Kaplan & Lin, 2000; Kaufman, 2009; Luthar & D’Avanzo, 1999; Ostrowsky & Messner, 2005). Kaufman’s (2009) longitudinal GST study found that depressive symptoms predicted suicidal thoughts, weekly drinking, running away, and violent offending among girls but “only” suicidal thoughts and running away among boys.

Ostrowsky and Messner’s (2005) GST study found victimized young adults were more likely to commit property and violent offenses than their nonvictimized counterparts, strains tended to have more impact on violent than property crimes, and strains were related to depression. Notably, strains were more commonly associated with depression among the young women than among the young men, but the young men who were strained and depressed were more likely to offend (Ostrowsky & Messner, 2005). A GST study of women and men involved in drug courts across the United States found recent sexual or physical abuse increased the risk of further substance use, and this was associated with (or mediated by) increased depression following either of these (sexual or physical) abuses (Zweig, Yahner, & Rossman, 2012). However, the resulting depression could not explain the reuse of substances for recent sexual abuse victims as completely as depression explained the drug reuse by recent physical abuse victims (Zweig et al., 2012).

Watts and Iratzoqui (2019) conducted one of the few GST studies that included girls and boys, ran the models separately for each, and included three types of abuse or maltreatment (i.e., physical, sexual, and neglect by a parent or guardian before sixth grade) and six self-reported offenses (i.e., violent, property, drug use, drug-selling, alcohol use, running away). They concluded that “child maltreatment increases delinquent behavior during middle adolescence, that different types of maltreatment differentially shape delinquent behavior, and that these relationships are marked more by gender similarity than gender difference” (p. 178). Specifically, their findings, summarized in Table 2.1, indicate that the three abuses impact both girls’ and boys’ violent offenses and running away the most, and alcohol and drug use the least. Child neglect is more frequently significantly related to the types of offenses than is physical or sexual abuse. Although other research shows that girls are significantly more likely than boys to be victimized by sexual abuse, this study found sexual abuse victimization is more likely to increase boys’ than girls’ subsequent offending—specifically, their violent offending, drug-selling, and running away. Indeed, all abuse in general impacts boys’ likelihood of subsequent offending far more often than it does girls’ likelihood in this study. One could speculate that the other GST research on gender differences in anger and depression (as reviewed earlier) could incite boys to be more likely to externalize (via anger) and girls to be more likely to internalize (via depression) their negative emotions from being abused and neglected.

Table 2.1

^aThe models controlled for race/ethnicity, parent’s education, public assistance (SES measure), self-control, peer deviancy, and closeness to mother. Offending variables are self-reported (not necessarily known by the criminal legal system). The abuse/maltreatment variables only measured these abuses before the start of sixth grade and if perpetrator was a parent or adult caregiver.

Source: Watts, S. J., & Iratzoqui, A. (2019). Gender, child maltreatment, and delinquency. Victims & Offenders, 14(2), 165–182.

Note: Data from U.S. National Longitudinal Study of Adolescent to Adult Health (Add Health). Sample = 14,322 youths followed over time.

Some GST research has included only girls and women in their samples. Belknap, Holsinger, and Little (2012) applied GST to incarcerated girls to study how different types of abuse (differentiated by family-perpetrated vs. non-family-perpetrated and sexual abuse vs. physical abuse) impacted girls’ self-harming (e.g., cutting, suicide attempts, etc.), while controlling for the girls’ sexual identities. Consistent with “community” (nonincarcerated sample) studies, they found that sexual minority status (SMS) girls (those who identified as lesbian or bisexual) reported more of all types of abuses (than their straight counterparts) and more self-harming behaviors than non-SMS (straight) girls. However, when controlling for abuse, the relationships between sexual identity (SMS vs. non-SMS) and self-harming disappeared (Belknap, Holsinger et al., 2012). Stated alternatively, the relationship between sexual identity and self-harming was indirect and completely explained by abuse; this indicated that SMS girls were disproportionately abused as a gender-based/homophobic response, and this was related to self-harming. (Thus, deterring homophobic assaults and bullying will likely deter the association between SMS and self-harming.) This needs to be tested in community samples as well.

A GST study of drug and alcohol use among women incarcerated in Oklahoma looked at a long list of strains (primarily different types of abuse and traumas), as well as anger, self-esteem, and antisocial behavior (Sharp, Peck, & Hartsfield, 2012). Consistent with GST, the greater the cumulative strain, the greater the women’s anger, and the more likely they were to abuse substances. Also, both sexual abuse and witnessing their mothers being abused were related to daily drug use. Contrary to GST, self-esteem and antisocial behavior were not related to daily drug or alcohol use (Sharp et al., 2012).