Criminal Justice Outcomes in Intimate and Non-intimate Partner Homicide Cases


3.5 Key independent variables: Intimacy, time, and gender


As noted above, of all social relationships, there may be none more intense than that of intimate partners because of the presence of sexual intimacy and physical proximity (Silverman & Kennedy, 1993). To capture the effect of intimacy, the first key independent variable is a dichotomous measure that distinguishes between intimate partner and non-intimate partner homicides.[17] The intimate partner category includes both current and former legal spouses, common-law partners and dating couples (i.e. boyfriends and girlfriends).[18] The non-intimate partner category includes family members (not including spouses), friends, acquaintances and strangers. In the Toronto sample, 20 percent of the cases involved killings between intimate partners (see Table 3.3 for coding and descriptive information for all independent and control variables).

Table 3.3: Description of Independent and Control Variables for Homicide Cases, Total Sample, Toronto, Ontario, 1974-2002

Key Independent Variables
Variable Description/Coding Mean (S.D.)
Victim-accused relationship Non-intimate partner = 0; Intimate partner = 1 .20 (.40)
Year case entered court Time period: 1974 to 1983 = 1; other = 0 .38 (.47)
Time period: 1984 to 1996 = 1; other = 0 .51 (.49)
Time period: 1997 to 2002 = 1; other = 0 .11 (.44)
Gender of accused Female = 0; Male = 1 .89 (.31)
Gender of victim Female = 0; Male = 1 .71 (.45)

Control Variables - Legal variables
Variable Description/Coding Mean (S.D.)
Criminal history of accused Accused had non-violent record (0, 1) .43 (.49)
Accused had violent record (0, 1) .13 (.34)
Role of accused in homicide Secondary = 0; Primary = 1 .85 (.36)
Number of accused One accused = 0; Multiple accused = 1 .16 (.37)
Number of victims One victim = 0; Multiple victims = 1 .04 (.20)
Case resolved at trial Guilty plea = 0; Trial = 0 .58 (.49)
Charge seriousness Interval (Least to most serious, 1 through 3) 2.28 (.62)
Conviction seriousness Interval (Least to most serious, 1 through 3) 2.7 (.89)

Control Variables - Characteristics of the accused
Variable Description/Coding Mean (S.D.)
Race/ethnicity of accused Non-white = 0; White = 1 .56 (.50)
Age of accused Age 18-24 (0, 1) .35 (.48)
Age 25-34 (0, 1) .36 (.48)
Age 35-44 (0, 1) .17 (.38)
Age 45-54 (0, 1) .08 (.27)
Age 55 and up (0, 1) .03 (.18)
Mean age of accused .31 (10.93)
Accused employment status Unemployed = 0; Employed = 1 .30 (.46)
Accused marital status Not married = 0; Married = 1 .41 (.49)
Accused psychiatric history No treatment = 0; Treatment = 1 .10 (.30)

Control Variables - Characteristics of the victim
Variable Description/Coding Mean (S.D.)
Victim’s ethnicity Non-white = 0; White = 1 .58 (.49)
Victim’s age Age newborn-17 (0, 1) .09 (.29)
Age 18-24 (0, 1) .19 (.39)
Age 25-34 (0, 1) .25 (.43)
Age 35-44 (0, 1) .22 (.41)
Age 45-54 (0, 1) .12 (.33)
Age 55 and up (0, 1) .13 (.33)
Mean age of victim .35 (16.12)
Victim’s employment status Unemployed = 0; Employed = 1 .40 (.49)
Victim’s marital status Not married = 0; Married = 1 .45 (.50)
Victim’s psychiatric history No treatment = 0; Treatment = 1 .06 (.23)
Victim’s criminal history No record = 0; Violent or non-violent record = 1 .30 (.46)

Control Variables - Characteristics of the incident
Variable Description/Coding Mean (S.D.)
Weapon use No gun used = 0; Gun used = 1 .24 (.43)
Location of killing Private = 0; Public = 1 .36 (.48)
Accused drinking/using drugs No alcohol/drugs = 0; Using alcohol/drugs = 1 .55 (.50)
Victim drinking/using drugs No alcohol/drugs = 0; Using alcohol/drugs = 1 .45 (.50)


Since the early 1970s, there has been enormous growth in the amount of public and professional attention given to violence within the family, but more particularly, to the problem of violence against women within intimate relationships. Work by feminists and grassroots organizations have drawn the attention of both members of the public and the legal profession to what was traditionally perceived to be a private family problem not appropriate for legal intervention (Dobash & Dobash, 1979; Schneider, 1994). As a result, in the past several decades, social and legal reforms have begun to target intimate violence as a public, rather than a private, concern. However, as noted above, there has been little systematic examination of the association between intimacy and criminal law over time. This represents a gap in the literature given that the amendments to or the implementation of various legislative policies in recent years may have lead to or represent changes in the attitudes of criminal justice officials and members of the public, generally, toward violence within intimate relationships. In other words, it is reasonable to expect that this heightened awareness of or sensitivity to intimate violence may influence the way in which criminal justice officials respond to some types of violent crime (Mitchell, 1991; Roberts, 1992). While this analysis cannot test for a direct relationship between changes in law and criminal justice responses to intimate violence, it is possible to examine whether such changes have occurred in tandem, documenting a possible association that can set the stage for future research. As a result, the second key independent variable captures the year in which a homicide case entered the court system, distinguishing between three separate time periods as described below.

The first key changes in legislation and/or policy were introduced in the early 1980s with important continent-wide reforms that began to transform the way in which the criminal justice system responded to cases of intimate partner violence. From 1983 to 1986, Offices of the Attorney General and the Solicitor General adopted policy directives that required police and Crown prosecutors to charge and prosecute all incidents of spousal abuse where there were reasonable and probable grounds to believe that an offence had been committed. These policies are often described as "pro-charging" and "pro-prosecution" policies; nonetheless, they are, in fact, the applicable standards for all criminal conduct. Their specific application, however, to cases of spousal abuse played a key role in highlighting the critical distinction between the treatment of spousal abuse by the criminal justice system as a "criminal matter" and its historical treatment of spousal abuse as a "private matter." Following the implementation of pro-charging policies, for example, police officers were required to make an arrest when there was "reasonable or probable grounds" to believe an offence had been committed. Prior to this, a more rigorous standard existed that stipulated an officer had to witness the offence or the resulting injuries before charges could be laid. This policy change had a significant effect on the number of common assault charges laid in incidents of spousal violence in this country and served as an impetus for change in both public and professional attitudes toward intimate violence. Additional legislative and/or policy changes targeting intimate violence occurred during the 1990s. The most relevant for the purposes of this study and to the treatment of intimate violence in the courts was Bill C-41 introduced in 1996. As noted above, Bill C-41 and the subsequent amendments (s.718.2 CCC) state that the abuse of a spouse or the abuse of a position of trust should be considered aggravating factors in sentencing an accused.

Based on the above, a three-category variable distinguishes among the distinct social and legal environments that were created by various changes in legislation and policy over a period of three decades. The first time period captures cases that entered the criminal justice system between and including 1974 to 1983 – the 10-year period prior to and including the introduction of pro-charging and pro-prosecution policies in Canada. The second 13-year time period – 1984 to 1996 – captures those cases that entered the court during the period that came after the implementation of pro-charging and pro-prosecution policies and up to the implementation of Bill C-41 in 1996. The final six-year period – 1997 to 2002 – captures those cases that came after the introduction of Bill C-41 and the related amendments to the criminal code. In the Toronto sample, Table 3.3 shows that 38 percent of the cases were dealt with during the early period of the study, 51 percent during the second period and 11 percent in the third or most recent period.[19]


While the role of intimacy in criminal law is the key focus of this study, intimacy and gender are intricately linked in crimes of interpersonal violence and so the separate and combined effects of the gender of the accused and the victim are also examined. A fairly persistent finding in sentencing research is that adult female offenders are treated more leniently than adult male offenders (see reviews, Bickle & Peterson, 1991; Daly & Bordt, 1995; Odubekun, 1992; Steffensmeier et al., 1993). In contrast, gender differences were found to be less common in research that looked at case dismissals and convictions (Nagel & Hagan, 1982). However, while previous research on criminal justice decision-making has controlled for gender of the accused, isolating the distinctive effects of intimacy requires that the gender of both the accused and the victim be taken into account (Felson et al., 1999). For example, females primarily victimize and are victimized by family members, especially male intimate partners, whereas males primarily victimize and are victimized by other males, strangers or otherwise (Browne & Williams, 1989; Reiss & Roth, 1993). In addition, the gender of the victim has also been shown to be associated with criminal justice outcomes in homicide (Gross & Mauro, 1989; Rapaport, 1991) and with the severity of conviction (Farrell & Swigert, 1986; Williams, 1976). To capture possible gender differences in punishment, a dummy variable measures whether the accused was male or female. Consistent with previous research on gender differences in violent crime, males represent the majority of those accused (89 percent) in the Toronto sample. A variable also captures whether the victim was male or female and, in the Toronto sample, slightly more than 70 percent of the victims were male.

Date modified: