Female-female criminal assault: an evolutionary perspective.

Fascination with male-male assault has been widespread in the social sciences. The archetypal "Saturday night brawl" has been described in terms of sequential structure (Felson 1982), its proximal triggers (Toch 1969), its relationship to alcohol consumption (Pernanen 1976), its inherent rule structure Fox 1977; Marsh, Rosser, and Harre 1978), its subcultural backdrop (Bernard 1990), and perhaps most strongly in terms of its interpersonal and symbolic functions (Felson and Tedeschi 1996; Katz 1988; Luckenbill 1977). Our information on female-female assault is by contrast rudimentary, reflecting our generally greater ignorance of the nature of assault by women.

Women are arrested for 13 percent of violent index crime in the United States. The female percentages of arrests in 1990 for specific violent crimes were 11 percent for homicide, 13 percent for aggravated assault, 15 percent for simple assault, 7 percent for carrying weapons, and 8 percent for robbery (Steffensmeier and Allan 1996). There has been relatively little change in these percentages over the past thirty years (Kruttschnitt 1993; Steffensmeier 1980; Steffensmeier and Cobb 1981). These figures are largely confirmed by data from victimization surveys: Women were reported to be the assailants in 14 percent of all violent crimes, 7 percent of robberies, 14 percent of aggravated assaults, and 17 percent of simple assaults (Kruttschnitt 1994). These figures are similar to those found in England and Wales. The female percentage of arrests is 15 percent of violent offenses and 7 percent of robberies. The British crime survey reports women to be the assailants in 12 percent of "contact" (violent) crime. They account for 2 percent of muggings, 6 percent of stranger assaults, 13 percent of acquaintance assaults, and 21 percent of domestic assaults (Home Office Statistical Bulletin 1996).

Self-report studies suggest that the sex gap may be smaller than that computed from official figures (see Chesney-Lind 1997; Kruttschnitt 1994; Smith and Visher 1980). However, self-report studies often use younger respondents and the acts may be of a less serious criminal nature. In the United States, the National Youth Survey obtained self-report data from 1,725 youths aged 11 to 17 (Canter 1982). They found that male youths exceeded female youths in all categories of violent offenses by a ratio of 3.5 to 1. For specific offenses, the percentages admitting to various offenses at least once in the preceding year were as follows: aggravated assault, 9 percent boys and 3 percent girls; hit student, 6 percent boys and 3 percent girts; hit teacher, 10 percent boys and 5 percent girls; hit parents, 6 percent of both sexes; gang fighting, 17 percent boys and 7 percent girls; and carrying a hidden weapon, 10 percent boys and 2 percent girls. This cohort was followed to the ages of between 21 and 27. In this age range, 8 percent of men and 1 percent of women admitted aggravated assault, 2 percent of men and no women admitted to gang fighting, 1 percent of each sex admitted to hitting a parent, 10 percent of men and 1 percent of women admitted to hitting someone at work, and 19 percent of men compared with 12 percent of women admitted to other target-unspecified assaults (Bureau of Justice Statistics 1989). A recent British study (Home Office Research and Statistics Department 1995) of 1,721 young people aged 14 to 25 reported current participation rates in offending for fighting (7 percent male youths and 3 percent female youths), beating nonfamily members (2 percent male youths and 0.5 percent female youths) and hurting with a weapon (2 percent male youths and 0.5 percent female youths).

The greater aggressiveness of men rather than women is confirmed by psychological studies using both laboratory and psychometric techniques. Meta-analyses of laboratory-based studies (Bettencourt and Miller 1996; Eagly and Steffen 1986; Knight, Fabes, and Higgins 1996) suggest that men display more physical aggression than do women with the effect size ranging from d = .30 to d = .91 and also more verbal aggression, although the effect size tends to be less marked (d =.05 to .46). A meta-analysis of questionnaire data (Archer 1997) on aggression also confirms a significantly higher score by men on physical aggression (d =.63 to.79) and verbal aggression (d =.33 to .39). Cross-cultural data indicate that men are everywhere rated as more aggressive than women (Williams and Best 1990). Anthropological data also concur that cross-culturally, men are more aggressive than women (Brown 1991; Ember 1981; Rohner 1976; Whiting and Edwards 1973).

The bulk of aggression by both men and women is directed at men (Bureau of Justice Statistics 1997; Daly and Wilson 1988; Eagly and Steffen 1986; Harris 1996; Kruttschnitt 1994). Women's aggression toward male cohabitants has received considerable attention. Studies using the Conflict Tactics Scale suggest that a similar percentage of women as men admit to moderate and serious violent acts in their relationship. Over 25 independent studies using the Conflict Tactics Scale have found about equal rates of male and female admissions to both minor and major acts of violence (Campbell 1993). Indeed, a number of studies have reported higher levels of female aggression in the cohabiting context (e.g., Archer and Ray 1989; Arias, Samios, and O'Leary 1987; McNeely and Mann 1990; O'Leary and Arias 1988; O'Leary et al. 1989; Riggs, O'Leary, and Breslin 1990; Stets and Straus 1989; Sugarman and Hotaling 1989). It is important to bear in mind, however, that when injuries rather than acts are considered, women are more seriously injured in domestic violence than men are. One of the few areas of female violence that has begun to receive careful scrutiny is partner homicide (Benedek 1982; Mann 1988; Ogle, Meier-Katkin, and Bernard 1995) consequent on the finding that although women commit only about 10 percent of all homicides, their percentage involvement in spousal homicide is much higher, varying between 19 percent and 55 percent (Kruttschnitt 1993). The typical offender is in her early 30s, a welfare recipient with an arrest record who kills a current or former husband or cohabitant after a series of domestic fights with the final episode often precipitated by the man's …