«ABSTRACT Younger women, relative to older women, incur elevated risk of uxoricide. Some evolutionary theorists attribute this pattern to men's ...»
REPRODUCTIVE AGE WOMEN ARE OVER-REPRESENTED
AMONG VICTIMS OF WIFE-KILLING
Todd K. Shackelford, Division of Science--Psychology, Florida Atlantic University,
2912 College Avenue, Davie, FL 33314
David M. Buss, Department of Psychology, University of Texas at Austin, Austin, TX 78712
Jay Peters, School of Social Work, University of Maine, Orono, ME 04469
ABSTRACTYounger women, relative to older women, incur elevated risk of uxoricide. Some evolutionary theorists attribute this pattern to men's evolved sexual proprietariness. Other evolutionary theorists propose an evolved homicide module for wife killing. An alternative to both explanations is that young women experience elevated uxoricide risk as a byproduct of marriage to younger men, who commit most acts of violence. We used 13,670 uxoricides to test these explanations. Findings show that (1) reproductive age women incur elevated risk of uxoricide; (2) younger men are over-represented among uxoricide perpetrators; and (3) younger women, even when married to older men, still incur excess risk of uxoricide.
Discussion examines competing explanations for uxoricide in light of these findings.
WIFE-KILLING AND WOMEN'S REPRODUCTIVE STATUS
The fact that men attempt to control their spouses using a variety of tactics (Buss, 1988; Buss & Shackelford, 1997) suggests a conflict of interest between the sexes. According to some evolutionary arguments, women have evolved sexual strategies, such as the desire to control their own sexuality, that are at odds with the strategies of their partners. One female strategy, for example, is to secure investment and commitment from one man while securing better genes from another man (Gangestad, 1993; Smith, 1984). The conflicts between the sexes that ensue can be called forms of "strategic interference," because the woman's strategy cannot be successfully enacted without interfering with the man's strategy, and viceversa (Buss, 1989).
According to Daly and Wilson (1988), Dickemann (1981), Buss (1988), and others, men have evolved several strategies designed to deter their spouses from committing adultery, ranging from vigilance to violence. At the most
level, men can control their wives by conferring benefits, by inflicting costs, or both. Not all men possess resources that might be used to confer benefits, and so these men are predicted to be especially prone to using violence, or the threat of violence, to control their partner's sexuality.
According to Daly and Wilson (1988), there is brinksmanship in using violence, and sometimes it slips over the edge: "Men...strive to control women...women struggle to resist coercion and to maintain their choices. There is brinksmanship and risk of disaster in any such contest, and homicides by spouses of either sex may be considered slips in this dangerous game" (p. 205). More recently, they note that "...the fatal outcome in [spouse killings] is hypothesized to be an epiphenomenal product of psychological processes that were selected for their nonlethal outcomes" (Wilson, Daly, & Daniele, 1995, p. 287). According to this argument, spousal homicide per se is not an adaptation, not a designed outcome, and does not imply that the killing of one's spouse ever was adaptive. Instead, homicide is an unintended outcome of the use of violence designed for control and deterrence, not designed for death.
An alternative evolutionary theory proposes that men have evolved specific homicide modules, including a spousal homicide module (Buss & Duntley, 1998). According to this theory, there have been some historical circumstances in which killing an unfaithful spouse, or one who has defected from the relationship, might have been adaptive. An infidelity by the wife might cause a man to devote 20 or more years of effort to another man's children, and the public discovery would inflict substantial reputational damage on him. Because evolution by selection operates on a relative metric, one man's loss of a wife is an intrasexual competitor's gain. According to Evolved Homicide Module Theory, although some instances of wife killing may be "slips" or "epiphenomena," most are intentional and designed products of the evolved homicide mechanisms.
The finding that younger wives tend to be killed more often than older wives poses a puzzle, especially for the "slip-up" theory of coercive control. Why would men be more inclined to kill younger women, since such women are higher in fertility and reproductive value than are older women? This finding is especially puzzling, on the Daly and Wilson (1988) slip-up theory, in that it defies the way in which men treat all other forms of "property." Men are not more likely to "destroy" a new, expensive car than an old, cheap car, so why would "male sexual proprietariness" lead men to kill younger wives more often than older wives? Daly and Wilson (1988, p. 206) offer the following explanation: "We propose that...men are most jealous of the youngest women [presumably because of their high reproductive value] and are therefore most inclined to behave coercively toward such wives... Paradoxically, the high homicide risk incurred by young wives is indicative not of their low worth from the male perspective, but of precisely the opposite."
Evolved Homicide Module Theory offers a different explanation (Buss & Duntley, 1998). Younger women are killed more often than older women because the damage to the husband inflicted by an infidelity or defection is commensurately greater and homicide is one way to reduce the damage. When an older woman of low reproductive value defects to a rival, little may be lost in the currency of reproduction. When a younger women defects to a rival, the husband's loss is compounded by the tremendous bonus that the rival gains. Thus, the greater risk of uxoricide experienced by younger women is consistent with Evolved Homicide Module Theory.
Both of these competing evolutionary theories may be challenged by a third explanation, which suggests that younger women incur greater risk of uxoricide not because they are so reproductively valuable, but rather as an incidental byproduct of the fact that young women happen to be married to young men. It is well known, for example, that young men between the ages of 16 and 24 commit the majority of acts of aggression, including homicide (Wilson & Daly, 1985). So it might be that the greater risk that young wives incur has nothing to do with their own age per se, but is a byproduct of a tendency of young men to use violence in general, combined with assortative mating for age which places young women into close proximity with dangerous young men. A key method for adjudicating among these competing theories is to determine whether the uxoricide rate for reproductive age women is greater than the uxoricide rate for post-reproductive age women, even after controlling for husband's age. This question can be addressed by comparing the uxoricide rate of reproductive age women with the uxoricide rate of post-reproductive age women across two groups: one in which the husband is older and one in which the husband is younger.
If the uxoricide rate for reproductive age women is higher than the uxoricide rate for post-reproductive age women among those women married to older husbands, this would provide strong evidence that reproductive age women are the special targets of uxoricide, and that this targeting is not attributable to the relatively youthful age of their husbands. This was the primary goal of the study.
METHODSDatabase The United States Federal Bureau of Investigation (FBI) requests information from each state on criminalhomicides. Supplementary Homicide Reports (SHRs) include incident-level data on every reported homicide, including the relationship of the victim to the offender, and the ages of the victim and offender.
The database analyzed for the present project includes SHRs for the years 1976-1994 (Fox, 1996), providing information on 429,729 homicides. Uxoricide rates were calculated according to relevant population estimates provided by the United States Census (available from the first author upon request).
Of the over 400,000 cases of homicide included in the database, 13,670 were cases in which a man killed the woman to whom he was legally married. All analyses were restricted to these cases (one case was omitted due to probable coding error: A three-year-old wife was murdered by her 31-year-old husband. Because of the large number of cases, the results do not change when this case is omitted). The average age of uxoricide victims was 39.4 years (SD = 15.4 years), ranging from 15 to 95 years. The average age of uxoricide perpetrators was 43.3 years (SD = 15.7 years), ranging from 16 to at least 98 years (ages 98 and older were coded in the database as 98 years; three such cases were coded for men).
Figure 1 shows uxoricides per million married women per annum as a function of the age of the murdered wife. The uxoricide rate is highest for teenage women who have the greatest reproductive value, as has been found in smaller samples (Daly & Wilson, 1988). The uxoricide rate for teenage women is about two times greater than that for women aged 20 to 24 years. The uxoricide rate for women 20 to 24 years is about 1.5 times greater than that for women aged 25 to 29 years. The clear trend is for the uxoricide rate to decrease with the reproductive value of the woman. Older, post-reproductive age women are killed by their husbands at much lower rates than are younger, reproductive age women. This trend shows a slight reversal at the oldest age category, for women who are 85 or older, a finding also reported by others and possibly reflecting "mercy killings" of wives with terminal illnesses.
We next investigated whether younger men were over-represented among uxoricide perpetrators.
Figure 2 shows uxoricides per million married men per annum as a function of the age of the uxoricidal husband. Consistent with previous work, relatively younger men inflict uxoricide at greater rates than do relatively older men. The highest uxoricide rate is for men in their early 20s. Paralleling the uxoricide victimization rates for women, the clear trend in uxoricide perpetration rates for men is a decrease with age, from the early 20s to the early 80s. Also consistent with previous work, the uxoricide rate appears to increase suddenly for men who are 85 years and older.
We next conducted a test to discover whether women married to much older men incur elevated risk of uxoricide. To facilitate future work on the relationship between uxoricide rate and spousal age discrepancy, we constructed Table 1. As far as we know, no previous work has presented detailed information about uxoricide rate as a function of the age discrepancy between spouses. For the present project, our interest was in comparing the uxoricide rates of women married to relatively older men with the uxoricide rates of women married to same-age men and relatively younger men. Women at the greatest risk of getting killed are under the age of 25 and married to men between 45 and 54 years (95.9 uxoricides per million per annum). Women who are toward the end of their reproductive years, between the ages of 35 and 44, and married to men in the 45 to 54 age bracket, incur only one-seventh the risk of being killed, with an annual rate of 13.6. Other age pairings show similar trends.
FIGURE 1. UXORICIDES PER MILLION MARRIED WOMEN PER ANNUM AS A
FUNCTION OF AGE OF MURDERED WIFE.Uxoricides per million married women per annum 20 25-29 35-39 45-54 65-74 85+ 20-24 30-34 40-44 55-64 75-84
FIGURE 2. UXORICIDES PER MILLION MARRIED MEN PER ANNUM AS A
FUNCTION OF AGE OF UXORICIDAL HUSBAND.Uxoricides per million married men per annum 20 25-29 35-39 45-54 65-74 85+ 20-24 30-34 40-44 55-64 75-84
TABLE 1. UXORICIDES PER MILLION MARRIED COUPLES PER ANNUM, BY
HUSBAND'S AGE AND WIFE'S AGE.___________________________________________________________________________
N/A 0.00 N/A 14.55 N/A 20.83 15.81 24.15 85+ ___________________________________________________________________________
Note. N/A = Population estimate of zero; therefore, uxoricide rate could not be computed.
Figure 3 is constructed from the data in Table 1 and shows uxoricides per million married couples per annum as a function of spousal age difference, in categories. In this figure, "1" indicates a one category difference between the age of a husband and the age of his wife, "2" indicates a two-category difference, and so on. Positive values refer to categorical differences in which a husband is older than his wife, whereas negative values refer to categorical differences in which a wife is older than her husband. "0" refers to cases
in which the husband and wife are in the same age category. The age categories are as follows, in years: