The Invisible Woman. Joanne Belknap

      4 Accounting for Gender–Crime Patterns

      The study of gender and crime has followed a trajectory, from finding that women offend differently, to exploring how gendered lives impact the nature and extent of offending.

      —Conover-Williams (2014, p. 449)

      Studies consistently show not only that girls and women commit far fewer crimes than boys and men but also that their offenses tend to be less serious and violent in nature. This chapter draws on prior and current research and data to describe girls and women’s offending and assess gender comparisons in offending. For a complete understanding of the studies and data, a gendered account of offending must be examined and understood in several contexts:

      The extent of offending is the frequency with which various offenses (e.g., homicide, burglary, arson) or combinations of offenses (e.g., delinquency, violent crimes, property crimes) are committed. The nature of offending addresses the type and seriousness of various offenses. Gender comparisons in offending are commonly evaluated by determining which offenses are gender-related and which are gender-neutral (Smart, 1976). Gender-related crimes are more likely to be committed by one sex/gender than the other. Gender-neutral crimes are equally likely to be committed by one sex/gender or the other. If an offense is gender-related, it is further identified by male- or female-gender-related to designate which gender is more likely to commit the crime. Most crimes are male-gender-related, and rape, homicide, and other violent crimes are especially so. The most common example of a female-gender–related crime is sex work (mostly measured as prostitution). There were no existing guidelines on what percentages constitute gender-neutral or gender-related until the fourth edition of this book, where I designated a crime as gender-related if there was a gender gap of more than 20 percentage points, and gender-neutral with a gender gap of less than 10 percentage points. Crimes with a gender gap between these—that is, greater than 10% and less than 20%—are designated as approaching gender-related (see Figure 4.1).

      Historically, legal codes for some offenses were written so that only one sex/gender could be a victim or offender. For example, until the 1970s many rape laws specified that only men (and boys) could be offenders and only women (and girls) could be victims. Similarly, many prostitution laws specified only women and girls as offenders (and no “victims,” but male clients were occasionally considered offenders). Significantly, even if the legal code is gender-neutral in identifying offenders and/or victims (i.e., not specifying “penis,” “woman,” etc.), the applications may still be gender-specific (Allison Morris, 1987). For example, even when the prostitution laws are changed to include male sex workers or to arrest clients as well as sex workers, police may be less likely to detect male sex workers because they do not envision them in these roles, and police, prosecutors and judges may continue to disregard sex worker clients as offenders (even if the law identifies them as law-breakers) because they do not fit their stereotypes of offenders (see Farley & Kelly, 2000). These responses are likely legacies of the positivist theorists (Chapter 2).

      Measuring Crime

      Crime rates are measured in many ways. In large national studies, typically, the rate is the number of offenses per 100,000 people in the population and measured by police arrests or some other official criminal legal system (CLS) response. The most popular data sets used to assess crime rates in the United States are the Uniform Crime Reports (UCR), yearly arrests reported by police departments across the country, which is gradually being replaced by the National Incident-Based Reporting System (NIBRS). (The most recent statistics are accessible through the Bureau of Justice Statistics website: http://www.ojp.usdoj.gov/bjs/.) The biggest shortcoming of UCR and NIBRS is they do not include crimes unknown and unreported to the police. NIBRS is preferable to the UCR in that it (1) includes more crimes (e.g., NIBRS includes domestic violence and UCR does not); (2) measures race/ethnicity far better; (3) provides far more detail about the crime incident, the victims and offenders (and how they are related to each other), victim injury, and property loss; and unlike UCR, (4) includes more arrest and clearance information (Pattavina, Carkin, & Tracy, 2017). However, NIBRS is not nationally representative data yet; currently it reports from around 7,000 law enforcement agencies covering around a third of the U.S. population in 36 states (A. Cooper, personal communication, July 10, 2019; Gavrilova, 2019).

      Description

      Figure 4.1 ● Assessing the Gender Gap in Offending Rates

      Note: For gender-related offenses, when the men/boys have the higher rate the offense is male-gender-related, and when the women/girls have the higher rate the offense is female-gender-related.

      But given that most crimes are never reported to the police, particularly those where victims feel stigmatized, which is most gender-based abuses, victimization surveys are typically preferable to UCR or NIBRS. In the United States this is primarily the National Crime Victimization Survey (NCVS, once called the National Crime Survey [NCS]), a survey of representative households from across the country collected by the federal government. Finally, some studies also use self-reported offending, which again, is more likely to reflect most crime rates. Also, self-reported offending is often compared with official records, and if matched by the individuals in both (self-report and offending), it is an ideal way to determine sexism, racism, and so on, in who is held more accountable for offending by the CLS. These victim or offender self-report measures of crime rates include distributing written or computer surveys and conducting one-on-one phone or in-person interviews. Clearly, if conducted repeatedly (such as the NCVS, UCR, NIBRS)—they can be used to examine changes over time.

      2009–2018 Arrest Rates From the UCR

      Although I warn against using solely UCR data, I use the most recent UCR data available at the time of writing this book for Table 4.1, because NIBRS, NCVS, and other data are far more difficult to access and NIBRS is not yet representative of the United States. Specifically, Table 4.1 presents 2018 UCR arrest data, compares offending from 2009 to 2018 (to examine potential changes in the gender gap for a 10-year period), and reports overall gender differences (all arrests) and gender differences solely for youth (juvenile arrests). The 2018 UCR data strongly support crime as almost exclusively male-gender-related: Boys and men were 73% of arrests, representing almost three times as many arrests as those of girls and women (27%) (see Table 4.1). Focusing solely on youth (under age 18), girls represent slightly more of the arrests (30%) than combining girls and women. Thus, the gender gap in arrests is slightly less (3%) among youths than among all ages combined.

      Table 4.1

      ^aThe 2009 rape figures are based on the legacy definition, and the 2018 rape figures are aggregate totals based on both the legacy and revised UCR definitions. For this reason, a percent change is not provided.

      ^b”Other