Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2018 Apr 27.
Published in final edited form as: Justice Q. 2016 Aug 17;34(4):652–673. doi: 10.1080/07418825.2016.1216153

The Competition-Violence Hypothesis: Sex, Marriage, and Male Aggression*

Patrick M Seffrin 1
PMCID: PMC5922783  NIHMSID: NIHMS912823  PMID: 29706687

Abstract

Sexually active men, who are not in a monogamous relationship, may be at a greater risk for violence than men who are sexually active within monogamous relationships and men who are not sexually active. The current study examines changes in sexual behavior and violence in adolescence to early adulthood. Data on male (n = 4,597) and female (n = 5,523) respondents were drawn from four waves of the National Longitudinal Study of Adolescent and Adult Health (Add Health). HLM regression models indicate that men who transition to a monogamous, or less competitive, mode of sexual behavior (fewer partners since last wave), reduce their risk for violence. The same results were not replicated for females. Further, results were not accounted for by marital status or other more readily accepted explanations of violence. Findings suggest that competition for sex be further examined as a potential cause of male violence.

Introduction

Since its publication, Sampson and Laub’s (1990) landmark study of previously incarcerated males has generated a flurry of research on marriage and crime. This research generally finds that after men marry they report less criminal behavior, are less likely to be arrested or incarcerated, and are less involved with drugs and deviant peer groups (see Blokland and Schipper, 2016 for a recent review). Marriage also demands of men that they no longer compete against one another for sexual partners, which historically has been a source of male aggression and continues to engender conflict (Campbell, 2005; Betzig, 1986; Buss, 2006; Keegan, 2004; van der Dennen, 1995).

The current study augments prior life course related theories (e.g., Laub and Sampson, 2003) by discussing sex, in addition to marriage, in relationship to violent behavior. While prior research shows that marriage may indeed reduce male criminal offending, there are at least two problems posed by focusing on marital unions in criminological research. The first problem is demographic. The median age of first-time marriage in the United States is about 26 for women and 28 for men (U.S. Census Bureau, 2010). Desistence from criminal behavior, on the other hand, typically begins in mid to late adolescence (Gottfredson and Hirschi, 1990). Early marriages may reduce involvement in crime and violence (e.g., Siennick et al., 2014; Theobald and Farrington, 2010); however, this subset of unions is not statistically representative of the youth population. Furthermore, cohabitation has come to replace marriage in recent decades as the type of union many young people engage in during their early adult years (Smock, 2000; Lesthaeghe and Neidert, 2006); yet rates of violent crime have declined despite this major social change (Zimring, 2006).

A second problem, related but not specific to the research on marriage, is that most criminological studies that have examined intimate relationships among adolescents and young adults have focused heavily on the qualities of a single union (Capaldi, Hyoun, and Owen, 2008; Eklund, Kerr, Stattin, 2010; Haynie et al., 2005; Giordano et al., 2010; McCarthy and Casey, 2008). Criminologists who conduct peer research by comparison rarely focus on the influence of a single friendship, even though these dyads tend to be longer lived than dating relationships (Giordano, 2003; Warr, 2002). While this singular focus makes sense in the context of marriage and perhaps cohabitation, it is far less applicable when studying youthful populations and patterns of sexual behavior. Over the course of adolescence and early adulthood, individuals will likely experience multiple intimate relationships. Relatively high levels of sexual exclusivity and emotional attachment may characterize some of these relationships, while others may be more fleeting as well as sexually non-exclusive (Manning, Longmore, and Giordano, 2005).

The perspective brought to bear on the data is grounded in research on evolved differences between the sexes (e.g., Darwin, 1871; Buss, 2003; Puts, 2010; Trivers, 1972; Zahavi, 1975). As a general rule of sexual behavior, also known as Bateman’s principle (1948), males often attempt to have sex frequently and with a variety of partners if given the opportunity, while females by comparison are more judicious in selecting partners and will reject sexual advances under a variety of circumstances, individual and cultural variation notwithstanding (Brown et al., 2009). A consequence of these divergent strategies is that males may need to compete against other males for a chance to be selected by a female. Sexual competition therefore holds the potential for violence, particularly when the romantic and sexual interests of one man are at odds with another.

Women are also competing for desirable partners, and whose choice of partners may play a significant role in the link between sex and violence. Men who show a willingness to take risks, have a high self-esteem, and a body that is physically imposing possess qualities that women may find desirable, but these qualities are also correlated with aggressive behavior (Apicella, 2014; Brewer and Howarth, 2012; Frederick and Haselton, 2007; Sell et al., 2009; Baumeister, Smart, and Boden, 1996). A broadly held preference among women for partners who embody these qualities may result in physically aggressive men being favored for sex. Thus because of female biases in mate selection, it follows that any observed link between men’s sexual behavior and their involvement in violence may be explained not as a direct result of men competing, but rather as a by-product of women choosing men on the basis of qualities which happen to correlate with physical aggression.

There is also the possibility that male competition and female choice conspire to create the link between sex and violence. If aggressive behavior in men is correlated with qualities that women find desirable, all that is needed for an aggressive competition to ensue is for young women to exercise a preference for desirable men, and for men to want sex. From this perspective both genders are culpable for the link between sexual behavior and violence—men for competing against each other for sexual partners and women for choosing the most desirable competitors. Monogamy offers men an escape from this unwitting conspiracy and is therefore central to the issues being explored in the current study.

The current study integrates prior research on sexual competition and violence with findings from criminology to advance a competition-violence hypothesis. It states that sexually active men, who are not in a monogamous relationship, are in competition against other men for access to females, and when challenged by a rival, denied access by a third party, or humiliated men may seek violent retribution. Women, by selecting partners on a non-random basis, play a central role in men’s competition for sex that inadvertently and disproportionately favors men who have a history of violence as well as the potential for being violent in the future. Transitioning to a monogamous, or less competitive, mode of sexual behavior may reduce men’s risk for violence. The hypothesis is tested in a longitudinal sample of males and females over a period of about twelve years. The current study looks to Thomas Hobbes’s (1651) principle causes of quarrel, routine activities theory (Cohen and Felson, R, 1979; Felson, M and Boba, 2009), as well as sexual selection theory (Darwin, 1871; Puts, 2010; Trivers, 1972; Zahavi, 1975) to expound on the competition-violence hypothesis.

Background

Prior evidence linking men’s sexual behavior to violence can be found in research that varies in method and approach. Historians and anthropologists note that competition for wives and conflict over adulterous behavior is chief among the reasons for primitive warfare in non-state and tribal societies (Betzig, 1986; Keegan, 2004; van der Dennen, 1995). While modern governments may no longer endorse sexual jealousy as a justification for war, men’s participation in violence often has a sexual motive, which is clearly present in cases of homicide throughout the industrialized world. Daly and Wilson (1988), for example, in a comprehensive analysis of homicide, conclude that sexual jealousy and competition among males are motivating factors in a large proportion of cases. Buss (2006) reports similar findings in research that asked a large sample of culturally diverse participants to describe scenarios in which they might want to kill someone (or may be killed). A reoccurring theme in the participants’ narratives often concerned sex and intimate relationships in which there was infidelity, a jilted lover, and the pursuit of justice to right these wrongs. Finally, research into the social behavior of men and women in bars and nightclubs confirms what was long suspected about violence in and around these places, that incidents are especially likely to occur when the romantic and sexual interests of different parties conflict (Graham et al., 2014, 2011, 2010).

The notion that one’s sexual and romantic interests may be linked to violence is also supported by research that employs experimental designs. Research conducted by Ariely and Loewenstein (2006) suggests that when men are aroused by the prospect of a sexual encounter they may behave more aggressively toward rival males as well as female partners, lending credibility to the notion that acts of violence are experienced by perpetrators as “crimes of passion” and seem to happen in the “heat of the moment”. Ainsworth and Maner (2012) present similar findings that males, but not females, are more likely to act out in an aggressive manner towards other males after being primed by researchers to think about sexual encounters. Furthermore, Thompson et al. (2006) finds that men, after being exposed to arginine vasopressin— a peptide thought to be linked to aggression and sexual arousal in men— exhibit facial expressions that are angrier and more threatening than controls and perceive less friendliness in the faces of other men. The women in this study, by contrast, reacted to the same peptide with friendlier facial expressions and perceived less animosity in the faces of other women. Collectively, these studies suggest that situations in which men’s sexual and romantic interests are at the forefront are the very situations in which violent behavior is likely to occur.

Male Competition

While violence and the situations in which it occurs are well documented, there are no established criminological theories for linking sexual competition to observed patterns of violence. Hobbes’s (1651) principle causes of quarrel continue to be widely discussed in the literature on violence (Pinker, 2011), and because of their general applicability, they serve as a useful starting place for unpacking the competition-violence hypothesis. Consider Hobbes’s first cause of quarrel, which is in fact competition. This first cause, applied to sexual competition, may be understood from a “routine activities” perspective on crime, which is focused on the interplay of motivated offenders, suitable targets, and capable guardians (Cohen and Felson, R. 1979; Felson, M. and Boba, 2009). Here the motivated offender is a male, who targets females that he believes will be receptive to his advances. He may be blocked, however, in pursing this goal by a guardian, who prevents his access, or by a rival male who has taken his place. If he persists despite resistance from third parties, violent confrontation may result (Parks et al., 2013).

Violence resulting from sexual competition may also be embedded within a longer timeline of events and may not be resolved in a single incident. As Buss’s (2006) research on homicide suggests, men may stew over a perceived indiscretion that kept them achieving their romantic or sexual goals, but then an opportunity arises to balance the scales of justice. In the mind of the aggrieved party, justice is achieved through violent retribution, but from the perspective of his rivals justice will not be done until this aggression is repaid in kind. This much is true of violence generally, which can produce vicious cycles of revenge, despite its initial causes being trivial and sometimes sexual in nature (Black, 1983; Harris, 2003; Kanazawa and Still, 2000; Katz, 1988; Wolfgang, Ferracuti, and Mannheim, 1967).

Hobbes’s two other principle causes of quarrel are fear and reputation. Reputation, or honor, is a re-occurring theme in the research on violence. Ethnographic descriptions of urban crime, studies on the differences in violent crime rates between northern and southern states, and historical analyses of violence all emphasize honor, and the fear of being dishonored or victimized, as a basis of male aggression (e.g., Anderson, 2000; Eisner, 2003; Elias, 1978; Nisbett and Cohen, 1996; Ruff, 2001). Fear and reputation may also play a role in the violence that is associated with sexual competition. The injury to honor that men may feel for failing to defend a current or potential partner against the sexual advances of rival men may give rise to retaliatory violence, to restore one’s reputation, as well as preemptive violence, for fear that the offense may happen again or simply to keep it from happening in the first place (Black, 1983; Felson, 1997; Mullins, Wright, and Jacobs, 2004).

Female Choice

While sexual competition may be important for explaining male aggression, female choice plays an integral role. Men who would use physical violence to gain a competitive advantage may possess other qualities that are sexually appealing to women, at least in the short-term. This much has been suggested in research by Rebellon and Manasse (2004) who found that highly delinquent males report relative success in attracting female dating partners. Rebellon and Manasse (2004) interpret these findings using a derivative of sexual selection theory known as the “handicap principle” (Zahavi, 1975). The handicap principle suggests that behaviors that are potentially costly to males—such as fighting and showing disregard for authority, but which are valued by females, perhaps for the strength and bravado they symbolize—will be implemented as tactics in male sexual competition (see also Palmer and Tilley 1995). Research on sexual selection theory also suggests that a preference for these qualities may have itself been selected for in females (Puts, 2010). This would help to explain why men have a penchant for violent behavior in the first place, in the sense that male aggression, and a preference for it among females, were selected for in the course of human prehistory.

Partnering with an aggressive and/or criminally involved male may have its advantages, especially in an unsafe environment where threats of violence are commonplace. Yet displays of dominance and physical aggression play just as well to an all male audience, who serve as a source of encouragement and validation, thereby reinforcing the behavior as well as its symbolic value in the peer culture (Messerschmidt, 1993). Furthermore, qualities in males that are correlated with aggressive behavior, such as confidence, bravado, and an athletic physique, may be attractive to women in the short-term, but in the long-term these are not necessarily the same qualities that women look for in a spouse or that would sustain a committed relationship (Buss, 2003).

Women’s choice of partners is further constrained by unfavorable economic conditions which can make men less desirable as marital partners (Wilson, 2011), and the sex ratio itself (Barber, 2009; Walsh, 2003), which may contain few men relative to the number of women, thus favoring men’s preference for frequent sex with a variety of partners, for which they may face competition from other men who are similarly inclined. And so without purpose or malicious intent, and while acting within the constraints of their environments, men and women conspire to make the pursuit of sex and romance a potentially violent affair.

Desistance

Desistance from sexual competition, and the violence that it produces, may require men to reassess definitions of what it means to act manly (Carlsson, 2013; Giordano, 2014; Rocque, Posick, and Paternoster, 2016; Messerschmidt,1993; Simons and Barr, 2014). Men, after all, do not pursue sexual opportunities endlessly as Batman’s principle might suggest. Men play more conventional heterosexual roles as well, such as husband, father, and reliable partner and provider. These role transformations however, like the pro-social effect of marriage itself, may depend on the level of commitment within relationships as well as a general departure from competitive modes of sexual behavior, which may prove difficult for men who feel that long-term relationships are inaccessible because of their poor economic status and/or inexperience with stable heterosexual unions (Parker and Reckenwald, 2008). Nevertheless, prior research shows that young males commit fewer delinquent acts while in a romantic relationship with a loved partner, but sex outside of relationships increases the risk of criminal offending (McCarthy and Casey, 2008). Further, males who do not begin a new relationship shortly after breaking up with a girlfriend increase their risk of future criminal offending (Laursen and Sweeten, 2012).

The break-up effect, noted previously, may highlight that newly single males are likely competing again for female partners, possibly in high risk places, such as bars, nightclubs, and parties, where social barriers are less defined and authority figures lacking (e.g., Leonard, Quigley, and Collins, 2002; Kim, 2013; Miller, 2012; Miller et al., 2005; Osgood and Anderson, 2004; Pridemore and Grubesic, 2011). Thus, changes in men’s sexual behavior may signal changes in their routine activities, and more importantly in their willingness to compete for female partners, thereby making them vulnerable to Hobbesian quarreling, selective female biases, and ultimately a greater risk of violent confrontation.

Current Study

The current study examines sexual behavior and violence in a longitudinal sample of males and females. The analyses are focused primarily on males, although results are also presented for females. The effects of sexual behavior are compared to other, more readily accepted, explanations of violence. Drug and alcohol use may increase the likelihood of violence, especially when used in social contexts such as parties and nightclubs, where relaxed social norms are paired with opportunities for males to make sexual advances (Graham et al., 2014, 2011, 2010). Childhood trauma, emotional attachment to caregivers, and local risk factors have also been shown to influence patterns of violence and risky sexual behavior (Harris, 2003; MacMillan 2001; Maimon and Browning, 2012; Widom, 2000). Retrospective reports of childhood physical and sexual abuse are controlled in the regression models, as are community levels of poverty and the proportion of unmarried males and females (Parker and Reckenwald, 2008; Porter and Purser, 2010). Emotional attachment to caregivers is controlled for in wave 1 of the data. Finally, marriage and full-time employment, which prior research suggests may deter men from criminal behavior, are controlled for in all waves (Laub and Sampson, 2003; Uggen, 2000).

The competition –violence hypothesis makes five statements that are tested in the analyses to follow:

  1. Changing to a monogamous, or less competitive, mode of sexual behavior (fewer partners since last wave), reduces violence.

  2. Changing to a more competitive mode of sexual behavior (multiple partners since last wave), increases violence.

  3. Men with a history of violence are selected for sex more often than their less aggressive peers.

  4. Both male competition and female choice contribute to the link between sex and violence.

  5. The influence of sexual behavior on violence cannot be accounted for by marital status or any other more readily accepted explanation of violent behavior.

Data

The National Study of Adolescent and Adult Health (Add Health) is the data source for the current study.1 The Add Health study is a nationally representative sample of adolescent youth containing four waves of longitudinal data. The sampling frame consisted of all high schools in the United States. A sample of 80 high schools and 52 middle schools was selected with unequal probability of selection. Respondents from grades 7 through 12 were initially surveyed during the 1994–1995 academic year. Waves 1 thru 4 of the restricted version of this dataset were used in the current study. The analytic sample includes males (n = 4,597) and females (n = 5,523) who participated in all four waves of the data. Fifty-four and one-half percent of the analytic sample is White/non-Hispanic; 19.5% African American/Black; 16% Hispanic/Latino; 7% Asian; 3% other.

Measures

Violence is measured with a 4-item index that asks respondents how often they have pulled a knife or gun on someone, seriously injured someone in a fight, gotten into a group fight, and shot or stabbed someone. These four items were selected because they appear in all waves of the data. Violence items were coded as “0” or “1” to eliminate differences in measurement and then summed to create a 5-point index, range 0–4 (see Table 1). While some of the violence items were originally coded dichotomously, with “yes” or “no” responses, other items were de facto dichotomous due to the high degree of right-skew in the distributions, and were re-coded as such. The violence index is therefore an approximate indicator of one’s ranking on an underlying continuum of violence potential, with high scores suggesting involvement in violence that is varied in type (some incidents more serious than others) and occurs with some regularity.

Table 1.

Descriptive Statistics for Males, Waves 1–4

Wave 1 Wave 2 Wave 3 Wave 4
Violence (0–4) 0.58 0.42 0.28 0.13
0.90 0.82 0.62 0.47
%Sexual Behavior (#partners since last wave)
 Highly Competitive (6 or more) 3.85 4.07 4.66 4.59
 Mid Competitive (4–5) 3.89 3.96 6.05 5.96
 Low Competitive (2–3) 9.72 8.42 21.84 19.73
 Monogamous (1) 7.72 7.16 42.46 53.36
 Non-Active (0) 74.81 76.4 24.99 17.36
Marriage and Employment
 %Married 0.41 0.41 12.99 42.77
 %Employed 3.57 8.22 52.95 63.02
Drug and Alcohol Use
 Drugs (0–4) 0.36 0.35 0.85 1.09
0.66 0.65 1.00 1.11
 Drunkenness (0–6) 0.63 0.79 1.40 1.26
1.28 1.44 1.59 1.45
Caregiver Abuse and Attachment
 Physical Abuse (0–5) 0.81
1.48
 Sexual Abuse (0–5) 0.70
0.43
 Attachment (1–5) 4.56
0.50
Local Risk Factors
 %Below Poverty Line 13.44
13.15
 %Never Married 24.80
8.50
Age 15.84 16.31 21.77 28.27
1.63 1.65 1.66 1.67
Months 0 10.99 76.26 154.39
- 1.78 2.87 3.49
N = 4,597
Open in a new tab

Note: Standard deviations appear in italics, beneath the means. Ranges appear in parentheses.

Sexual behavior is measured in terms of different modes, with each mode represented by a dichotomous or discrete variable. In waves 2–4 of the data, respondents were asked how many sexual partners they had since the last interview. In wave 1, the number of lifetime sex partners was recorded. Respondents reporting zero sex partners represent the sexually non-active mode. Virgins were included in the non-active mode. Respondents with one sex partner represent the monogamous mode, and respondents with more than one sex partner represent the competitive mode. The competitive mode is further decomposed into categories of 2–3 partners (low competitive), 4–5 partners (mid competitive), and 6 or more partners (highly competitive). The highly competitive mode makes up a small but consistent percentage of the sample in each wave (3–4 percent), however large enough to be treated as a discrete variable. Sexual behavior is therefore represented with variables that are mutually exclusive and exhaustive of the sex partner variable. Furthermore, while the number of sexual partners reported is an imperfect proxy for degrees of competitiveness, the concept is similar to dating/mating effort, or the extent to which individuals express interest and involvement in dating and sex, which is shown in prior research to be predictive of delinquent behavior (Rowe et al., 1997; Seffrin et al., 2009).

Marriage is measured with a dichotomous indicator. In waves 3 and 4 of the data, respondents indicated the month and year in which they were married. Respondents were coded as being married in waves 1–4, provided they were married for at least twelve months at the time of the interview. In each wave, respondents were recorded as employed if they indicated current and fulltime employment, defined as working 35 hours or more in a week.

Drug use is measured with a 4-item index that asks respondents how often, since the last interview, have they used marijuana, cocaine, injection drugs, and hallucinogens or other drugs. In wave 1, lifetime drug use was recorded. These four items were selected because they appear in all waves of the data. Drug items were coded as “0” or “1” to eliminate differences in measurement and then summed to create a 5-point index, range 0–4. Similar to the violence items, several of the drug items were de facto dichotomous due to the high degree of right-skew in the distribution, and were re-coded as such. Alcohol use “drunkenness” is measured with a single item which ranges from 0 “never” to 6 “almost every day” and asks respondents how often in the past 12 months they got drunk. This item appears in all waves of the data.

Childhood abuse is measured with two items that appear in wave 3 of the data. These items refer to experiences that occurred prior to the 6th grade. Respondents were asked how often a parent or caregiver hit, kicked, or slapped them (i.e., physical abuse), and how often they were forced to engage in sexual relations with a parent or caregiver (i.e., sexual abuse). Each item ranges 0 “never” to 5 “more than ten times.” Attachment to caregivers is measured with a 6-item scale, recorded in wave 1 of the data. Respondents were asked if they feel that their mother, father, caregiver are warm and loving, cares about them, and how close they are to mother, father, and/or other caregiver. Each item ranges 1 “not at all” to 5 “very much.” Cronbach’s alpha for the scale is 0.81.

Local risk factors are measured with two items that appear in wave 1 of the data. The percent of residents living below the poverty line and percent of never married adults were recorded in the census-block groups where respondents were living shortly before the time of the wave 1 interview.

Age is measured in years, and months is measured as the number of months since the wave 1 interview. The longitudinal design of the current study requires that time-varying observations are demarcated in terms of the length of time that has elapsed between waves. Thus at wave 1, months equals “0” for all respondents, but then varies between respondents in the three subsequent follow-up interviews, depending on the respondents’ respective interview dates. Further, the months variable was logged before being entered into the regression analyses (Singer and Willet, 2003).

Analytic Strategy

Descriptive statistics are presented for the male sample in Table 1. Results from longitudinal HLM regression models are presented in Tables 2 and 3. Fixed and random effects are used to statistically control for the influence of individual characteristics that are relatively stable over time but which may be correlated with variations in violence and sexual behavior (e.g., Gottfredson and Hirschi’s conception of ‘low self-control’).2 Within-individual and between-individual effects for sexual behavior are presented in Tables 2 and 3, along with control variables. Within-individual components are created by subtracting each respondent’s average on X (the between-individual average) from each time-ordered observation of X. This technique allows the within-individual components to be interpreted as fixed effect regression estimates, as change in the dependent variable caused by a change in an independent variable (Raudenbush and Bryk, 2002). Once within-individual effects are statistically controlled, the between-individual effects for sexual competitiveness suggest the degree to which violent males are successful at attracting sexual partners, or likewise, the degree to which non-violence is associated with monogamous commitment. However, given that women usually choose who they have sex with, a statistically significant link between violence and male sexual behavior would suggest a bias among females for sexual partners who happen to be relatively aggressive. If both within-individual and between-individual effects for sexual behavior are statistically significant, this would suggest that male competition, as well as female choice, contribute to the link between sex and violence.

Table 2.

Longitudinal HLM Regression Analysis of Sex and Violence

Fixed Effects b S.E. b S.E.
Sexual Behavior (Within-Individual)
 Highly Competitive Comparison group
 Mid Competitive −0.159*** 0.033 −0.140*** 0.033
 Low Competitive −0.232*** 0.028 −0.203*** 0.028
 Monogamous −0.312*** 0.027 −0.249*** 0.028
 Non-Active −0.363*** 0.027 −0.294*** 0.027
Sexual Behavior (Between-Individual)
 Highly Competitive Comparison group
 Mid Competitive −0.389*** 0.074 −0.382*** 0.072
 Low Competitive −0.611*** 0.059 −0.555*** 0.058
 Monogamous −0.857*** 0.055 −0.722*** 0.055
 Non-Active −0.987*** 0.052 −0.782*** 0.053
Marriage and Employment (Within-Individual)
 Married −0.201** 0.071
 Married × Months 0.049** 0.016
 Employed −0.014 0.013
Drug and Alcohol Use (Within-Individual)
 Drug Use 0.052*** 0.008
 Drunkenness 0.058*** 0.004
Caregiver Abuse and Attachment
 Physical Abuse 0.018*** 0.004
 Sexual Abuse 0.049*** 0.013
 Attachment 0.001 0.012
Local Risk Factors
 %Below Poverty Line 0.151** 0.050
 %Never Married 0.112 0.078
Age and Time
 Age −0.017*** 0.002 −0.021*** 0.002
 Months −0.054*** 0.004 −0.054*** 0.005
Random Variance Components
 Within-Individual 0.014*** 0.001 0.014*** 0.001
 Between-Individual 0.479*** 0.016 0.434*** 0.015
R-square
0.125 0.171
BIC
36279.4 35458.9
N = 4,597
Open in a new tab

Intercepts were estimated but are not shown. Race/ethnicity and the between-individual components for marriage, employment, drug and alcohol use are included in model 2, but not shown in the table.

p.*<.05,

**

<.01,

***

<.001.

Table 3.

HLM Analysis of Sex and Violence: Controlling for Gender and Sexual Status at Wave 1

Males Females
Virgin Wave 1 Sexually Active Wave 1 Virgin Wave 1 Sexually Active Wave 1
N = 3,439 N = 1,158 N = 4,506 N = 1,017
b S.E. b S.E. b S.E. b S.E.
Sexual Behavior (Within-Individual)
 Highly Competitive Comparison group
 Mid Competitive −0.107* 0.044 −0.156** 0.055 −0.071 0.040 0.003 0.050
 Low Competitive −0.197*** 0.037 −0.180*** 0.047 −0.070* 0.034 −0.022 0.043
 Monogamous −0.237*** 0.036 −0.215*** 0.048 −0.066* 0.033 −0.059 0.043
 Non-Active −0.244*** 0.036 −0.226*** 0.050 −0.045 0.033 −0.036 0.043
Sexual Behavior (Between-Individual)
 Highly Competitive Comparison group
 Mid Competitive −0.424*** 0.103 −0.333** 0.115 −0.261** 0.083 −0.157 0.103
 Low Competitive −0.536*** 0.083 −0.552*** 0.093 −0.192** 0.070 −0.106 0.086
 Monogamous −0.734*** 0.079 −0.689*** 0.090 −0.264*** 0.068 −0.166* 0.084
 Non-Active −0.744*** 0.079 −0.728*** 0.104 −0.257*** 0.068 −0.092 0.091
Marriage (Within-Individual)
 Married −0.188* 0.075 −0.075 0.166 −0.018 0.043 0.178 0.113
 Married × Months 0.046** 0.017 0.023 0.038 0.007 0.010 −0.038 0.026
R-square 0.110 0.214 0.081 0.115
N = 10,120 (total sample)
Open in a new tab

Intercepts were estimated but are not shown. Models control for all variables present in Table 2. p.

*

<.05,

**

<.01,

***

<.001.

Results

Table 1 shows the distribution of modes of male sexual behavior in each wave of the data. Participation in the highly competitive mode (6 or more partners) remains fairly constant, with roughly 4 percent of males involved in this mode over the course of the study. The same goes for mid-level competitiveness (4–5 partners), whereas participation in low levels of competitiveness (2–3 partners) rises from about 9 percent in wave 1 to 18 percent in wave 4. Monogamous involvement (1 partner) also increases over time, with about 8 percent monogamous in wave 1 and 53 percent in wave 4. Participation in the non-active mode (0 partners) decreases over time, with about 75 percent non-active in wave 1, down to 17.4 percent in wave 4.

Table 2 presents a longitudinal HLM regression analysis that includes both between-individual and within-individual effects for the different modes of sexual behavior. Model 1 compares the highly competitive mode to all other modes of sexual behavior. Findings suggest that when boys and young men transition from a highly competitive mode of sexual behavior to one that is monogamous violence declines by 0.312 points on a 5-point index. Transitioning from monogamy to a highly competitive mode increases violence by the same amount (result not shown). Violence also declines when men transition from a highly competitive mode to one that less competitive or non-active.

Model 2 controls for marital status and employment, drug and alcohol use, childhood abuse and attachment to caregivers, poverty, and the proportion of never married adults in the respondents’ census-block group. The interaction term (marriage × months) is also included in the model to account for the time-sensitive effect of marriage found in prior research (e.g., Theobald and Farrington, 2010). Marriage is shown here to reduce violence (b = −0.201), but according to the positive and significant interaction term ‘marriage × months’ (b = 0.049) its effect on violence diminishes the later men wait to marry. Employment status is not statistically significant. Drug use, drunkenness, and childhood abuse are all statistically significant whereas attachment to caregivers is not. Local levels of poverty are significant, whereas percent never married is not. All together these more readily accepted explanations of violence explain only about five additional percentage points in the overall variation in violence. Further, the effect of monogamy is reduced in size, but remains statistically significant, net of these controls.3

The significant between-individual effects indicate that men with little history of violence are more likely to be monogamous or non-active than competitive, whereas men with extensive histories of violent behavior are more likely to be highly competitive than monogamous or non-active. Both within-individual and between-individual effects are statistically significant, which suggests that female choice also plays a role in linking sex to violence. More specifically, the models suggest that women are more likely to have sex with men who have been violent in the past, and may be violent in the future, than with men who are less physically aggressive.

Post-Hoc Analysis

Post-hoc analyses were conducted to examine whether and to what degree the research hypotheses apply to women and how prior sexual experience may influence the link between sex and violence. Table 3 shows results for male and female respondents who were already sexually active at the time of the wave 1 interview (i.e., at least one lifetime sexual partner reported), as well as results for respondents who were sexual virgins at the time of wave 1. The former of these two groups constitutes a subset of respondents who are sexually precocious compared to their peers, which according to prior research may indicate an increased risk for delinquent behavior (e.g., Downing and Bellis, 2009). Results for virgin males and males who were sexually active at wave 1 are similar to each other and to the results shown in Table 2. However for sexually active females the within-individual effects are not statistically significant. Marriage is non-significant for precocious males, and non-significant for females in both models.

When the sample is limited to virgin females, changes in sexual behavior are found to have a significant relationship with violence, which may suggest that peer networks and routine activities have also changed to include hanging out with boys and partying (e.g., McGloin and DiPietro, 2013). The more precocious females, on the other hand, may have already made these transitions and may not compete as males do for sexual partners. Further, the significant between-individual effects for virgin females suggest that physically aggressive women have more sexual partners, on average, than less aggressive women, which is similar to what was found for men, however the same findings do not hold consistent for sexually precocious females.

Discussion of Findings

Sex is a defining feature of intimate relationships as well as a source of violent conflict, yet prior research on criminal desistance has ignored the influence of sexual behavior, focusing instead on marriage and employment to explain changes in criminal offending. Findings from the current study indicate that males, who transition to a less competitive mode of sexual behavior as evidenced by a reduction in sex partners from previous waves, reduced their risk for violent behavior. Changes in sexual behavior were shown to be more consistent and stronger in predicting violence than marriage and employment. The pro-social effect of marriage was not explained by changes in sexual behavior however, which suggests that marriage reduces male aggression for reasons other than monogamous commitment (e.g., Laub and Sampson, 2003). Findings also indicate that violent men tend to have more sexual partners than men who are less aggressive, which suggests that female choice, in addition to male competition, plays a role in linking sexual behavior to violence. The same findings were not replicated for females. This suggests that female competition does not carry the same potential for violence as male competition, if for no other reason than the fact that men are competing against other men, and women against women (Felson, 1996). As prior research suggests, in a competition for desirable partners, the malicious actions that transpire between women are more likely to take the form of psychological rather than physical aggression, such as manipulation and character assassination (Archer, 2004; Salmivalli and Kaukiainen, 2004; Vaillancourt, 2013). Furthermore, the results of the analyses were drawn from longitudinal data, which controlled for changes in drug and alcohol use as well as early childhood factors that may have accounted for the sex-violence link, but did not. Thus, the findings of this study are in support of the research hypotheses and suggest that competition for sex be further examined as a potential cause of male violence.

Limitations and Directions for Future Research

The current study is limited in several respects, foremost by the ambiguity introduced when attempting to interpret changes in sexual behavior. While changes in sexual behavior were shown to have a statistically significant influence on changes in violence, the data cannot say for sure that sexual competition is the reason why. For a male to accumulate a relatively high number of sexual partners in a short time suggests competition—fighting and wrangling with other men—although the same outcome could be achieved with little to no violence. Further, a growing body of research suggests that, rather than marriage and monogamous unions affecting violence, the relationship may also work the other way around, with crime and antisocial behavior affecting union formation and stability (Barnes et al., 2014; Bersani and Doherty, 2013; Lyngstad and Skardhamar, 2013). Unfortunately, detailed information on specific relationships was limited to a small sample of sexual partners, making it difficult to discern why some of the men became monogamous while others did not. More research is needed here on the conditions that push men and women in and out of committed relationships as it may relate to involvement in crime and violence.

A second set of limitations concerns the treatment of sexual behavior as an independent influence on violence, although prior research suggests that both behaviors may be influenced by the same social conditions. Men with a history of violence often have unstable and short-term relationships with women and come disproportionately from segments of society that have been marginalized by poverty and racial segregation (Wilson, 2011). Although marriage and monogamous relationships may offer men a healthier lifestyle, these are long-term and costly investments, whose formation and stability may therefore depend on life being predictable and free of violence (Daly and Wilson, 2005). Furthermore, high crime rates and staggering levels of unemployment can make marriage and monogamous relationships inaccessible to men who would otherwise benefit from a stable heterosexual union. A more inclusive theory of sex and violence will need to pay special attention to high-risk and marginalized groups of men and women who live under conditions of legal and economic dysfunction.

The current study was also focused primarily on male behavior; however women are not passive agents in the arena of sexual competition. Campbell (1995) for example explains that women may become violent when defending a male partner from the sexual advances of a rival female. Women can also be the victims of competitive male aggression, especially if they are believed to have been unfaithful to a partner (Archer, 2014; Capaldi et al., 2012). Further, the hypothesis that females prefer males who possess qualities which happen to correlate with aggressive behavior was tested through indirect means by examining the between-individual effects of male sexual behavior on violence. A more direct test of this hypothesis would require an assessment of women’s preferences in men as well as the situations in which they may be likely to partner with an aggressive male. Finally, women’s risk of criminal involvement is disproportionately affected by male associates (McGloin and DiPietro, 2013). Co-offending occurs frequently among youthful offenders, and for women in particular, criminal involvement often involves male co-offenders (Warr, 2002). The female gender and the composition of peer networks should be considered more thoroughly in future research on crime and intimate relationships.

The violence analyzed in the current study is self-reported, along with most all of the other measures, and is almost certainly non-lethal. However incidents of non-lethal violence are relatively high in the same places and within the same socio-demographic groups that have high homicide rates and official crime statistics are generally correlated with self-reports (Gottfredson and Hirschi, 1990; Zimring, 2006). Thus, the competition-violence hypothesis, by focusing on sex in addition to marriage, suggests a broader scope of behavior for criminologists to consider in future research on violence and criminal desistance.

Footnotes

*

This research uses data from Add Health, a program project directed by Kathleen Mullan Harris and designed by J. Richard Udry, Peter S. Bearman, and Kathleen Mullan Harris at the University of North Carolina at Chapel Hill, and funded by grant P01-HD31921 from the Eunice Kennedy Shriver National Institute of Child Health and Human Development, with cooperative funding from 23 other federal agencies and foundations. Special acknowledgment is due Ronald R. Rindfuss and Barbara Entwisle for assistance in the original design. Information on how to obtain the Add Health data files is available on the Add Health website (http://www.cpc.unc.edu/addhealth). No direct support was received from grant P01-HD31921 for this analysis.

1

Harris, K.M., C.T. Halpern, E. Whitsel, J. Hussey, J. Tabor, P. Entzel, and J.R. Udry. 2009. The National Longitudinal Study of Adolescent to Adult Health: Research Design [WWW document]. URL: http://www.cpc.unc.edu/projects/addhealth/design. The sampling weight variables provided by Add Health were not used in this study due to the use of male and female subsamples and the implementation of hierarchical linear models that account for the complex error structure in the data. Missing cases for certain variables constituted less than two percent of the analytic sample. The mean was imputed for the census-block variables, percent below the poverty line and percent never married, while zero was imputed for all other variables with missing cases and was interpreted as ‘no evidence for the behavior in question’

2

The multilevel/mixed regression is a two-level model, expressed below as a composite of level 1 and level 2.

Composite Model:

Yij=∑i=1n0iγkXik+(εij+ζ0i+ζ1iTimeij).

The notation (εij + ζ0i + ζ1iTime ij) is the stochastic part of the model and represents the random variance components (ζ) for time, the intercept, and the error (ε), at level 1. The notation ∑i=1n0iγkXik is the structural part of the model and represents a vector of fixed effects and the intercept for the ith respondent at the jth time of measurement, at level 2 (see Singer and Willet, 2003).

3

The results from the HLM models were compared to results from a GEE logit model (general estimating equation), fit for repeated observations as well as to a general linear model for estimating fixed effects. The “genmod” SAS procedure produced the GEE model estimates. The SAS “general linear model” procedure was used to estimate the fixed effects regression parameters. The SAS “mixed” procedure was used to estimate the HLM regression parameters shown in Tables 2 and 3. Results from these alternative models were nearly identical to that of the HLM models.

References

  1. Ainsworth SE, Maner JK. Sex begets violence: Mating motives, social dominance, and physical aggression in men. Journal of Personality and Social Psychology. 2012;103(5):819. doi: 10.1037/a0029428. [DOI] [PubMed] [Google Scholar]
  2. Anderson E. Code of the street: Decency, violence, and the moral life of the inner city. WW Norton & Company; 2000. [Google Scholar]
  3. Apicella CL. Upper-body strength predicts hunting reputation and reproductive success in Hadza hunter–gatherers. Evolution and Human Behavior. 2014;35(6):508–518. [Google Scholar]
  4. Archer J. Can evolutionary principles explain patterns of family violence? Psychological Bulletin. 2013;139(2):403. doi: 10.1037/a0029114. [DOI] [PubMed] [Google Scholar]
  5. Archer J. Sex differences in aggression in real-world settings: A meta-analytic review. Review of General Psychology. 2004;8(4):291. [Google Scholar]
  6. Ariely D, Loewenstein G. The heat of the moment: The effect of sexual arousal on sexual decision making. Journal of Behavioral Decision Making. 2006;19(2):87. [Google Scholar]
  7. Barber N. Countries with fewer males have more violent crime: Marriage markets and mating aggression. Aggressive behavior. 2009;35(1):49–56. doi: 10.1002/ab.20291. [DOI] [PubMed] [Google Scholar]
  8. Barnes JC, Golden K, Mancini C, Boutwell BB, Beaver KM, Diamond B. Marriage and involvement in crime: A consideration of reciprocal effects in a nationally representative sample. Justice Quarterly. 2014;31(2):229–256. [Google Scholar]
  9. Bateman AJ. Intra-sexual selection in Drosophila. Heredity. 1948;2(3):349–68. doi: 10.1038/hdy.1948.21. [DOI] [PubMed] [Google Scholar]
  10. Baumeister RF, Smart L, Boden JM. Relation of threatened egotism to violence and aggression: The dark side of high self-esteem. Psychological Review. 1996;103(1):5. doi: 10.1037/0033-295x.103.1.5. [DOI] [PubMed] [Google Scholar]
  11. Bersani BE, Doherty EE. When the ties that bind unwind: Examining the enduring and situational processes of change behind the marriage effect. Criminology. 2013;51(2):399–433. [Google Scholar]
  12. Betzig LL. Despotism and differential reproduction: A Darwinian view of history. Transaction Publishers; 1986. [Google Scholar]
  13. Black D. Crime as social control. American Sociological Review. 1983:34–45. [Google Scholar]
  14. Blokland A, De Schipper N. How important are life-course transitions in explaining desistance? Global Perspectives on Desistance: Reviewing what We Know and Looking to the Future. 2016:144. [Google Scholar]
  15. Brown GR, Laland KN, Mulder MB. Bateman’s principles and human sex roles. Trends in Ecology & Evolution. 2009;24(6):297–304. doi: 10.1016/j.tree.2009.02.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Brewer G, Howarth S. Sport, attractiveness and aggression. Personality and Individual Differences. 2012;53(5):640–643. [Google Scholar]
  17. Buss DM. The murderer next door: Why the mind is designed to kill. Penguin; 2006. [Google Scholar]
  18. Buss DM. The evolution of desire: Strategies of human mating. Basic books; 2003. [Google Scholar]
  19. Campbell A. Aggression. In: Buss DM, editor. The handbook of evolutionary psychology. Hoboken, NJ: John Wiley & Sons; 2005. pp. 628–652. [Google Scholar]
  20. Campbell A. A few good men: Evolutionary psychology and female adolescent aggression. Ethology and Sociobiology. 1995;16(2):99–123. [Google Scholar]
  21. Capaldi DM, Knoble NB, Shortt JW, Kim HK. A systematic review of risk factors for intimate partner violence. Partner Abuse. 2012;3(2):231. doi: 10.1891/1946-6560.3.2.231. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Capaldi DM, Kim HK, Owen LD. Romantic partners’ influence on men’s likelihood of arrest in early adulthood*. Criminology. 2008;46(2):267–299. doi: 10.1111/j.1745-9125.2008.00110.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Carlsson C. Masculinities, persistence, and desistance. Criminology. 2013;51(3):661–693. [Google Scholar]
  24. Cohen LE, Felson M. Social change and crime rate trends: A routine activity approach. American sociological review. 1979:588–608. [Google Scholar]
  25. Daly M, Wilson M. Carpe diem: Adaptation and devaluing the future. Quarterly Review of Biology. 2005;80(1):55–60. doi: 10.1086/431025. [DOI] [PubMed] [Google Scholar]
  26. Daly M, Wilson M. Homicide. Transaction Publishers; 1988. [Google Scholar]
  27. Darwin C. Sexual selection and the descent of man. Murray; London: 1871. [Google Scholar]
  28. Downing J, Bellis MA. Early pubertal onset and its relationship with sexual risk taking, substance use and anti-social behaviour: A preliminary cross-sectional study. BMC Public Health. 2009;9(1):1. doi: 10.1186/1471-2458-9-446. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Eisner M. Long-term historical trends in violent crime. Crime and Justice. 2003:83–142. [Google Scholar]
  30. Elias N. The Civilizing Process. 1978;1 and 2 [Google Scholar]
  31. Eklund JM, Kerr M, Stattin H. Romantic relationships and delinquent behaviour in adolescence: The moderating role of delinquency propensity. Journal of Adolescence. 2010;33(3):377–386. doi: 10.1016/j.adolescence.2009.09.002. [DOI] [PubMed] [Google Scholar]
  32. Felson RB. Anger, aggression, and violence in love triangles. Violence and victims. 1997;12(4):345–362. [PubMed] [Google Scholar]
  33. Felson RB. Big people hit little people: Sex differences in physical power and interpersonal violence. Criminology. 1996;34(3):433–452. [Google Scholar]
  34. Felson M, Boba RL, editors. Crime and everyday life. Sage; 2009. [Google Scholar]
  35. Frederick DA, Haselton MG. Why is muscularity sexy? Tests of the fitness indicator hypothesis. Personality and Social Psychology Bulletin. 2007 doi: 10.1177/0146167207303022. [DOI] [PubMed] [Google Scholar]
  36. Giordano PC, Lonardo RA, Manning WD, Longmore MA. Adolescent romance and delinquency: a further exploration of Hirschi’s “cold and brittle” relationships hypothesis*. Criminology. 2010;48(4):919–946. doi: 10.1111/j.1745-9125.2010.00208.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Giordano PC. Effective Interventions in the Lives of Criminal Offenders. Springer; New York: 2014. Gender, crime, and Desistance: Toward a theory of cognitive transformation; pp. 41–62. [Google Scholar]
  38. Giordano PC. Relationships in adolescence. Annual Review of Sociology. 2003:257–281. [Google Scholar]
  39. Gottfredson MR, Hirschi T. A general theory of crime. Stanford University Press; 1990. [Google Scholar]
  40. Graham K, Bernards S, Wayne Osgood D, Abbey A, Parks M, Flynn A, Wells S. “Blurred lines?” Sexual aggression and barroom culture. Alcoholism: Clinical and Experimental Research. 2014;38(5):1416–1424. doi: 10.1111/acer.12356. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Graham K, Bernards S, Wells S, Osgood DW, Abbey A, Felson RB, Saltz RF. Behavioural indicators of motives for barroom aggression: Implications for preventing bar violence. Drug and Alcohol Review. 2011;30(5):554–563. doi: 10.1111/j.1465-3362.2010.00252.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Graham K, Wells S, Bernards S, Dennison S. “Yes, I do but not with you”: Qualitative analyses of sexual/romantic overture-related aggression in bars and clubs. Contemporary drug problems. 2010;37(2):197–240. [PMC free article] [PubMed] [Google Scholar]
  43. Harris CR. A review of sex differences in sexual jealousy, including self-report data, psychophysiological responses, interpersonal violence, and morbid jealousy. Personality and Social Psychology Review. 2003;7(2):102–128. doi: 10.1207/S15327957PSPR0702_102-128. [DOI] [PubMed] [Google Scholar]
  44. Haynie DL, Giordano PC, Manning WD, Longmore MA. Adolescent romantic relationships and delinquency involvement. Criminology. 2005;43(1):177–210. [Google Scholar]
  45. Hobbes T. Leviathan or The Matter, Forme and Power of a Commonwealth Ecclesiasticall and Civil. 1651. [Google Scholar]
  46. Kanazawa S, Still MC. Why men commit crimes (and why they desist) Sociological Theory. 2000;18(3):434–447. [Google Scholar]
  47. Katz J. Seductions of crime: Moral and sensual attractions in doing evil. Basic Books; 2008. [Google Scholar]
  48. Keegan J. A history of warfare. Random House; 2011. [Google Scholar]
  49. Kim RH. Age-sensitive effect of adolescent dating experience on delinquency and substance use. Crime & Delinquency. 2013 0011128713486072. [Google Scholar]
  50. Larson M, Sweeten G. Breaking up is hard to do: Romantic dissolution, offending, and substance use during the transition to adulthood*. Criminology. 2012;50(3):605–636. [Google Scholar]
  51. Laub JH, Sampson RJ. Shared beginnings, divergent lives: Delinquent boys to age 70. Harvard University Press; 2003. [Google Scholar]
  52. Leonard KE, Quigley BM, Collins RL. Physical aggression in the lives of young adults: Prevalence, location, and severity among college and community samples. Journal of interpersonal violence. 2002;17(5):533–550. [Google Scholar]
  53. Lesthaeghe RJ, Neidert L. The second demographic transition in the United States: Exception or textbook example? Population and Development Review. 2006;32(4):669–698. [Google Scholar]
  54. Lyngstad TH, Skardhamar T. Changes in criminal offending around the time of marriage. Journal of research in crime and delinquency. 2013;50(4):608–615. [Google Scholar]
  55. Maimon D, Browning CR. Adolescents’ Violent Victimization in the Neighbourhood Situational and Contextual Determinants. British Journal of Criminology. 2012;52(4):808–833. [Google Scholar]
  56. Manning WD, Longmore MA, Giordano PC. Adolescents’ involvement in non-romantic sexual activity. Social Science Research. 2005;34(2):384–407. [Google Scholar]
  57. Macmillan R. Violence and the life course: The consequences of victimization for personal and social development. Annual Review of Sociology. 2001:1–22. [Google Scholar]
  58. McCarthy B, Casey T. Love, sex, and crime: Adolescent romantic relationships and offending. American Sociological Review. 2008;73(6):944–969. [Google Scholar]
  59. McGloin JM, DiPietro S. Girls, friends and delinquency. Oxford University Press; New York, NY, USA: 2013. pp. 313–330. [Google Scholar]
  60. Messerschmidt JW. Masculinities and crime: Critique and reconceptualization of theory. Rowman & Littlefield Publishers; 1993. [Google Scholar]
  61. Miller J. Individual offending, routine activities, and activity settings: Revisiting the routine activity theory of general deviance. Journal of Research in Crime and Delinquency. 2012 0022427811432641. [Google Scholar]
  62. Nisbett RE, Cohen D. Culture of honor: The psychology of violence in the south. Westview Press; 1996. [Google Scholar]
  63. Osgood DW, Anderson AL. Unstructured socializing and rates of delinquency. Criminology. 2004;42:519. [Google Scholar]
  64. Palmer CT, Tilley CF. Sexual access to females as a motivation for joining gangs: An evolutionary approach. Journal of Sex Research. 1995;32(3):213–217. [Google Scholar]
  65. Parker KF, Reckdenwald A. Concentrated disadvantage, traditional male role models, and African American juvenile violence*. Criminology. 2008;46(3):711–735. [Google Scholar]
  66. Parks MJ, Osgood DW, Felson RB, Wells S, Graham K. Third party involvement in barroom conflicts. Aggressive behavior. 2013;39(4):257–268. doi: 10.1002/ab.21475. [DOI] [PMC free article] [PubMed] [Google Scholar]
  67. Puts DA. Beauty and the beast: Mechanisms of sexual selection in humans. Evolution and Human Behavior. 2010;31(3):157–175. [Google Scholar]
  68. Pinker S. The better angels of our nature: Why violence has declined. New York: Viking; 2011. [Google Scholar]
  69. Porter JR, Purser CW. Social disorganization, marriage, and reported crime: A spatial econometrics examination of family formation and criminal offending. Journal of Criminal Justice. 2010;38(5):942–950. [Google Scholar]
  70. Pridemore WA, Grubesic TH. Alcohol outlets and community levels of interpersonal violence: Spatial density, outlet type, and seriousness of assault. Journal of Research in Crime and Delinquency. 2011 0022427810397952. [Google Scholar]
  71. Raudenbush SW, Bryk AS. Hierarchical linear models: Applications and data analysis methods. Vol. 1. Sage; 2002. [Google Scholar]
  72. Rebellon CJ, Manasse M. Do “bad boys” really get the girls? Delinquency as a cause and consequence of dating behavior among adolescents. Justice Quarterly. 2004;21(2):355–389. [Google Scholar]
  73. Rocque M, Posick C, Paternoster R. Identities through time: An exploration of identity change as a cause of desistance. JQ: Justice Quarterly. 2016;33(1):45–72. doi: 10.1080/07418825.2014.894111. [DOI] [Google Scholar]
  74. Rowe DC, Vazsonyi AT, Figueredo AJ. Mating-effort in adolescence: A conditional or alternative strategy. Personality and Individual Differences. 1997;23(1):105–115. [Google Scholar]
  75. Ruff JR. Violence in early modern Europe 1500–1800. Cambridge University Press; 2001. [Google Scholar]
  76. Sampson RJ, Laub JH. Crime and deviance over the life course: The salience of adult social bonds. American Sociological Review. 1990:609–627. [Google Scholar]
  77. Salmivalli C, Kaukiainen A. “Female aggression” revisited: Variable-and person-centered approaches to studying gender differences in different types of aggression. Aggressive behavior. 2004;30(2):158–163. [Google Scholar]
  78. Siennick SE, Staff J, Osgood DW, Schulenberg JE, Bachman JG, VanEseltine M. Partnership transitions and antisocial behavior in young adulthood a within-person, multi-cohort analysis. Journal of Research in Crime and Delinquency. 2014;51(6):735–758. doi: 10.1177/0022427814529977. [DOI] [PMC free article] [PubMed] [Google Scholar]
  79. Simons RL, Barr AB. Shifting perspectives: Cognitive changes mediate the impact of romantic relationships on desistance from crime. Justice Quarterly. 2014;31(5):793–821. doi: 10.1080/07418825.2012.704388. [DOI] [PMC free article] [PubMed] [Google Scholar]
  80. Skardhamar T, Monsbakken CW, Lyngstad TH. Crime and the Transition to Marriage the Role of the Spouse’s Criminal Involvement. British Journal of Criminology. 2014;54(3):411–427. [Google Scholar]
  81. Seffrin PM, Giordano PC, Manning WD, Longmore MA. The influence of dating relationships on friendship networks, identity development, and delinquency. Justice Quarterly. 2009;26(2):238–267. doi: 10.1080/07418820802245052. [DOI] [PMC free article] [PubMed] [Google Scholar]
  82. Sell A, Cosmides L, Tooby J, Sznycer D, von Rueden C, Gurven M. Human adaptations for the visual assessment of strength and fighting ability from the body and face. Proceedings of the Royal Society of London B: Biological Sciences. 2009;276(1656):575–584. doi: 10.1098/rspb.2008.1177. [DOI] [PMC free article] [PubMed] [Google Scholar]
  83. Singer JD, Willett JB. Applied longitudinal data analysis: Modeling change and event occurrence. Oxford university press; 2003. [Google Scholar]
  84. Smock PJ. Cohabitation in the United States: An appraisal of research themes, findings, and implications. Annual Review of Sociology. 2000:1–20. [Google Scholar]
  85. Theobald D, Farrington DP. Why do the crime-reducing effects of marriage vary with age? British journal of criminology. 2010:azq060. [Google Scholar]
  86. Thompson RR, George K, Walton JC, Orr SP, Benson J. Sex-specific influences of vasopressin on human social communication. Proceedings of the National Academy of Sciences. 2006;103(20):7889–7894. doi: 10.1073/pnas.0600406103. [DOI] [PMC free article] [PubMed] [Google Scholar]
  87. Trivers RL. Parental investment and sexual selection. In: Campbell B, editor. Sexual Selection and the Descent of Man. Aldinc; Chicago: 1972. pp. 136–179. [Google Scholar]
  88. Uggen C. Work as a turning point in the life course of criminals: A duration model of age, employment, and recidivism. American Sociological Review. 2000:529–546. [Google Scholar]
  89. U.S. Census Bureau, Decennial Censuses, 1890 to 1940, and Current Population Survey, Annual Social and Economic Supplements, 1947 to 2015.
  90. Vaillancourt T. Do human females use indirect aggression as an intrasexual competition strategy? Philosophical Transactions of the Royal Society of London B: Biological Sciences. 2013;368(1631):20130080. doi: 10.1098/rstb.2013.0080. [DOI] [PMC free article] [PubMed] [Google Scholar]
  91. van der Dennen JMG. The origin of war: The evolution of a male-coalitional reproductive strategy. 1 & 2. Origin Press; 1995. [Google Scholar]
  92. Walsh A. The sex ratio: A biosocial explanation for racial variation in crime rates. In: Walsh A, Ellis L, editors. Biosocial Criminology: Challenging environmentalism’s supremacy. Hampshire, NY: Nova Science; 2003. pp. 61–82. [Google Scholar]
  93. Warr M. Companions in crime: The social aspects of criminal conduct. Cambridge University Press; 2002. [Google Scholar]
  94. Widom CS. Childhood victimization: Early adversity, later psychopathology. National Institute of Justice Journal. 2000;242:3–9. [Google Scholar]
  95. Wilson WJ. When work disappears: The world of the new urban poor. Vintage: 2011. [Google Scholar]
  96. Wolfgang ME, Ferracuti F, Mannheim H. The subculture of violence: Towards an integrated theory in criminology. Vol. 16. London: Tavistock Publications; 1967. [Google Scholar]
  97. Zahavi A. Mate selection—a selection for a handicap. Journal of Theoretical Biology. 1975;53(1):205–214. doi: 10.1016/0022-5193(75)90111-3. [DOI] [PubMed] [Google Scholar]
  98. Zimring FE. The great American crime decline. Oxford University Press; 2006. [Google Scholar]

RESOURCES