The Invisible Woman. Joanne Belknap
were more likely than boys to report running to a friend’s house when they ran away, hurting their assault victims when the victims were students, and using force for reasons other than to get money. Regarding the context of committing assaults, boys were more likely than girls to report being on drugs during the assault, hurting their victims if their victims were not students, beating their victims or attacking them with a weapon, and having their victims cut or hospitalized. Thus, there were few gender differences in the context of offending for status, vandalism (property), or theft offenses, and the context of offending resulted in gender gaps more often in the commission of violent offenses.
Intersections With Race/Ethnicity and Class
Race/ethnicity is vital to understanding crime rates because of the potential for it to be skewed as a result of racist discrimination and how race/ethnicity are related to class and wealth (currently and through wealth legacies) and access to legal employment. Kruttschnitt (1996) points out that ignoring racial variations in gender comparisons of offending is “short-sighted” because it presumes that all races “are similarly situated in social life and so influenced by the same risk factors” (p. 131). In 1981, D. K. Lewis instructed that it is almost impossible to make these important distinctions about racial/ethnic variations among women (and among men) because of how the UCR and prison data are typically reported. Unfortunately, the problem still exists. For example, if one accesses the Bureau of Justice Statistics website to examine crime rates for a specific year or over time, the tables typically report looking at one, two, and possibly three variables at a time, but the format makes it impossible to understand how, for example, the rates break down when controlling for gender/sex, race/ethnicity, and age at the same time. Fortunately, government reports sometimes do this for the public, but these breakdowns are not always accessible. Additionally, until recently the racial/ethnic breakdowns were usually restricted to African American and white, which fails to account for variations among Latinx, Asian American, Indigenous, and other racial/ethnic individuals.
Significantly, official statistics such as the UCR and NIBRS, court convictions, and imprisonment rates rely on CLS actors’ perceptions of whether a crime occurred. Theoretically, police arrest, prosecutors prosecute, and judges and juries convict only when they believe there is a strong likelihood that a person committed a crime. Thus, a major threat to the validity of using official statistics of crime rates is that discrimination distorts the statistics. For example, if Latinas are more likely than white women to be arrested for the same offense, the official statistics would exaggerate the Latinas’ offending relative to white women. If African American women are more likely than Latinas to be arrested for the same offense, then official statistics will inflate African American women’s crime rates. Thus, when differences are found based on gender, race/ethnicity, class, age, and so on, it is not clear whether there are distinct actual offending differences in the categories or whether the differences among the categories represent differential (discriminatory) processing of offenders (e.g., sexist, racist, classist, etc. practices).
D. K. Lewis’s (1981) review of research found that “correctional” statistics suggest that African American women “display somewhat greater involvement with violent and other personal crimes than white women” (p. 93). But Lewis also cautions that when examining arrest records or other data to compare various races’ offending patterns, it is important to control for age because the higher ratio of African American to white females’ violent crime rates might partially be explained by the fact that the African American population as a whole is younger (and thus in a higher-risk age group) than the white population. On the other hand, the age difference between African Americans and whites cannot by itself, according to Lewis, explain the offending differences between white and African American women; these analyses must also consider economic deprivation, gender status inequality, socialization, gender role expectations, and racism.
Black women, then, display gender role behavior, a social status, and a crime pattern, all of which contradict acceptable feminine behavior as defined by the dominant society. They tend to be assertive, function as unmarried heads of household, and be convicted for violent person crimes. In short, they epitomize the type of deviant women the criminal legal system is committed to punish (D. K. Lewis, 1981, p. 102).
D. K. Lewis’s (1981) work, although reported in the early 1980s, is unfortunately still relevant today. Richie (2012) describes that a significant aspect of racist and classist criminalization has to do with cultural stereotypes that extend less harsh reviews of white, middle-class women and girls’ behaviors and how the media, government, feminist movement, Black communities, and academics have all failed to protest the criminalization of victimized Black women and girls. A government report of youths arrested in the United States in 2003 indicates that the overrepresentation of African Americans was not gendered; they were equally overrepresented among girls and boys (H. N. Snyder & Sickmund, 2006). The far-reaching effects of racism and how it intersects with sexism can be best understood by Richie’s (2006) description of risk factors for substance abuse, and how substance abuse by poor women of Color is responded to by the CLS:
Institutional neglect is also indicated as a major risk factor, and given racial and class hierarchy, it follows that women of color from low-income communities are most likely to experience the negative consequences of substance abuse; their treatment needs are not addressed and they are sanctioned by the criminal legal system most harshly. (Richie, 2006, p. 139)
It is also important to recognize that social class affects one’s opportunities and can affect treatment by individuals in society in general (e.g., teachers, neighbors, store owners, etc.) and by the police, courts, and so on. Class affects the likelihood that one will turn to crime for survival. Class also impacts a charged person’s ability to afford bail and hire an attorney, practices highly related to conviction and sentencing results. Unfortunately, data gatherers in both official statistics and many self-report studies are either unwilling or unable to account for class differences, so less is documented about this powerful variable.
One self-report study on inner-city and suburban youths (arguably a class measure) found suburban girls were the most likely to have tried alcohol and the most likely to have tried marijuana, whereas inner-city boys were the least likely to have tried alcohol and inner-city females were the least likely to have tried marijuana (Luthar & D’Avanzo, 1999). Contrary to popular class and race stereotypes, suburban males were most likely to have tried hard drugs, followed distantly by inner-city boys, inner-city girls, and suburban girls (respectively). A more recent study points to young people of high-income white families as the most likely to enter adulthood as “secret deviants” who do the most partying but are the least likely to be caught or arrested by officials, whereas African American youths were far more likely to be arrested for similar drinking and drug use behaviors (Hagan & Foster, 2006). Among girls and boys, Whites reported more drinking and drug-using than Latinx and African Americans. Across genders, youths from the wealthiest 25% of the study partied the most, and this was higher for the wealthiest white boys than the wealthiest white girls; youths from the poorest 25% partied the least (Hagan & Foster, 2006).
Studies that simultaneously control for race, gender, class, and age often find these are important predictors of offending behavior, as well as treatment by the CLS. Clearly, making simple racial, class, or gender comparisons is less useful than examining their intersections. It is important to emphasize the danger in relying solely on official CLS statistics (e.g., police and court records), in that they may be a better reflection of bias than offending. Researchers should also use self-report surveys and interviews to present a more valid measure of offending.
Zimmerman and Messner’s (2010) evaluation of youth self-reported violence in Chicago from 1994 to 2002 found gender convergence as the level of neighborhood disadvantage increased. It is useful to note, though, that police presence is likely much higher in these more disadvantaged neighborhoods, which could explain at least some of this relationship. Zimmerman and Messner also found that youth in more disadvantaged neighborhoods were more likely to have violent peers, and although boys had more violent peers than girls, peer violence appeared to impact the girls more than the boys. They also found that peer intimacy modified the relationship between peer violence and self-reported violent behavior, which they claim helped explain the greater effect of