The Impact of Law Reform on the Processing of Sexual Assault Cases
Gunn, Rita, Linden, Rick, The Canadian Review of Sociology and Anthropology
In her analysis of the history of rape laws, Christine Boyle (1984) has shown how these laws reflected a significant conflict in values. Society strongly condemned sexual attacks upon women, particularly those perceived as chaste and virtuous, and the laws provided harsh penalties for offenders. At the same time, legislators (virtually all of whom were male) seemed to fear innocent men being convicted and provided the accused with extraordinary protection against false charges of rape that many felt women were likely to make (Boyle, 1984).
Critics had identified many problems with the law (Clark and Lewis, 1977; DeKeseredy and Hinch, 1991; Hinch, 1996; Los, 1994), among them the criminal justice system's seeming greater concern for the complainant's credibility than for the innocence or guilt of the accused, which allowed degrading and stressful examination of a victim's previous sexual history. The law applied procedural and evidentiary standards to sexual offences such as corroboration and recent complaint, which were not applied to other offences. Rape was defined as a crime that could not apply to a spouse.
These critics felt the law reinforced stereotypes of rape that have long hampered efforts to prosecute and convict rapists. These stereotypes include myths such as: a woman cannot be raped unless she wants to be; evidence of injury is necessary to prove that a woman really resisted her attacker; women who "lead men on" are themselves responsible for the outcome; women who engage in risky behaviour such as hitchhiking or drinking alone in bars deserve to be raped; and, a woman who has had prior sexual relations with a man or with his friends has implicitly consented to further intercourse even if she says no.
Bill C-127: Legislative Reform
In response to these problems, the Parliament of Canada proclaimed Bill C-127 on January 4, 1983.(1) Included in this new legislation were significant changes such as:
* the reclassification of the crime of "rape" to three levels of sexual assault based on aggravating factors;
* disqualification of evidence concerning the complainant's background that is not pertinent to the case (rape shield law);
* removal of spousal immunity;
* making sexual assault gender-neutral; and
* changing the rules of evidence concerning consent, corroboration and recent complaint.
This change in legislation was intended to redefine sexual assault as a crime of violence rather than as a sexual act, and to reduce the harassment and degradation of victims by the justice system. Proponents hoped the new legislation would encourage more victims to report sexual assaults, ensure that victims were treated more fairly, and increase the conviction rate for sexual assaults, thereby enhancing the deterrent effect of the law.
Research on the Impact of Rape Law Reform
Several researchers have evaluated the impact of C-127. Using national police statistics, Renner and Sahjpaul (1986) compared the treatment of sexual assaults from 1973 to 1982 with post-legislation, 1983 data. The number of reported sexual assaults increased significantly in 1983, while the number of most other types of crime declined. There was no change in the percentage of cases reported as "founded," and small statistically significant increases in the percentage cleared and percentage charged. Roberts (1990) looked at national rates of reported sexual assaults before and after the legislation. He found that the number of complaints of sexual assault remained relatively stable in the six years prior to the implementation of Bill C-127, but that it increased by 127% in the six years after implementation. There was also a relative decrease in the severity of offences, which suggests that there has been a move away from the former emphasis on penetration (Department of Justice, 1990). However, there was no increase in either the proportion of cases classified as founded (Roberts, 1990) or in conviction rates (Department of Justice, 1990). Roberts and Gebotys (1993) conclude that the legislation
was successful in achieving one of the goals that inspired it. Ironically, its success has been in attracting more victims into the system, rather than in changing the way that the system responds to complaints of criminal sexual aggression (1993: 168).
The Department of Justice carried out a study of the impact of the legislation in six cities - Vancouver, Lethbridge, Winnipeg, Hamilton, Montreal and Fredericton - Saint John (Clark and Hepworth, 1994). While there was considerable variation from one city to the other, the findings were consistent with those of other studies. There was virtually no change in the proportion of cases reported as founded or in rates of clearance by charge. Conviction rates were higher in some cities, but lower in others: the net impact was therefore minimal.
Researchers have also assessed the impact of rape reform legislation in the United States. Among these have been studies in Michigan (Marsh et al., 1982), Washington (Loh, 1980), California (Polk, 1985; LeBeau, 1987); and a recent comparative study done in six U.S. jurisdictions (Spohn and Horney, 1992). The legislative changes evaluated in these studies were similar to those made in Canada, although the Canadian reforms were broader than those in most states. Effects reported in the American studies were modest at best, suggesting that the legal reform had a very limited impact on the operation of the criminal justice system.
The study carried out by Spohn and Horney is worthy of note because of its scope. Their comparison of the impact of rape law reform in six different states allowed for comparative analysis among states with different degrees of reform. Legislative change ranged from reforms enacted in Michigan and Illinois, which were similar in scope to those introduced in Canada, to those in Georgia and Washington, D.C., which made some changes relating to corroboration and prior sexual history, but that, in both cases, still required evidence of genital penetration and some evidence of resistance. The study used data covering 15 years to show long-term trends. The effect of "history" on the dependent variables was controlled to some extent by the fact that different jurisdictions had different intervention points (different dates of legislative change) while "history" factors such as an increased social concern for rape victims were likely to have been relatively constant from one jurisdiction to another.
Spohn and Horney conducted a time-series analysis using the date of legal change as the independent variable and several dependent variables, including offences reported to the police, number of indictments and number of convictions. Their results showed that the reforms had little effect. Only one jurisdiction showed changes that could be attributed to the legislative reform. That jurisdiction (Detroit) had significant increases in reports and in the ratio of indictments to reports. However, there was no increase in the proportion of cases resulting in conviction or in the likelihood of incarceration. Thus, with one partial exception, the effects of rape law reform in the United States appeared to be minimal.
In this paper, we present the results of an evaluation of the impact of Bill C-127 on the processing of sexual assault cases through the criminal justice system in Winnipeg, Manitoba. This research was part of the multi-city study discussed by Clark and Hepworth (1994). Other research using national police statistics has shown that the law had a limited impact on charges and convictions. The use of police and court file data in this study allowed us to see if there were changes that were not evident in the national data. Because the law was clearly intended to change the way in which sexual assault cases were treated by the courts, we used logistic regression to examine the impact of several independent variables on charges and convictions before and after the new legislation. Logistic regression allowed us to determine if the rape reform legislation led to changes in the way in which these variables affect charge and conviction. For example, we have been able to determine whether variables reflecting the character of the victim have diminished in importance following the implementation of Bill C-127.
For this evaluation, information was taken from police records, Crown files and court documents. The study years consisted of the two years prior to the enactment of Bill C-127 (1981-1982) and the two years after (1984-1985). The two study periods were compared to analyse the extent to which any changes occurred that might be attributed to the implementation of the new legislation.
a) Police Files Police reports signify the first official stage of a sexual assault complaint. It is at this level that a case is deemed to be "founded" or "unfounded." A classification of unfounded suggests that the officer receiving the complaint does not believe the complainant. This may be due to lack of evidence, or to other factors, which are often discretionary (e.g., the officer believes the complainant is too intoxicated to recall details; there are contradictions in the complaint; the complainant has a history of unfounded complaints). The officers handling the complaint decide whether they think the complaint is credible. Charging also occurs at this level, although advice may be sought from the Crown on the advisability of proceeding with a case.
The Winnipeg police received approximately 740 sexual assault complaints in 1981-1982 and 1180 in 1984-1985. These totals included all sexual offences, including incest and other offences not relevant to this evaluation. A 50% random sample of sexual offences was drawn manually (every second card) from the card files of the Bureau of Police Records.
From the random sample, cases which were not relevant to the evaluation were selected out. The 1981-1982 sample included the crimes of rape, attempted rape, indecent assault, sexual intercourse with a female between 14 and 16, sexual intercourse with a female under 14, gross indecency and buggery. To permit a comparison with the post-reform sexual assault data, we excluded incest cases except when combined with any of the above offences. The 1984-1985 sample was drawn from all cases that involved a complaint classified as sexual assault (Criminal Code, 246.1), sexual assault with a weapon, threats or causing bodily harm (Criminal Code, 246.2) and aggravated sexual assault (Criminal Code, 246.3). As in the pre-reform sample, incest was included only when combined with one of the three levels of sexual assault. The 1981-1982 sample consisted of 315 incidents involving 317 complainants and 306 accused persons. The 1984-1985 sample consisted of 523 incidents involving 528 complainants and 469 accused persons.
Incidents that were one-time events and involved more than one complainant or more than one accused were classified as a single incident of sexual assault. Ongoing offences that had multiple complainants and involved one accused or, alternatively, multiple accused and one complainant were classified as separate incidents. Because multiple accused and/or multiple complainants were involved in several cases, the total number of complainants and accused differs from the total number of incidents.
Information obtained from police files included sociodemographic details of the complainants and accused persons, description of the incidents (time, location, resistance, injuries sustained, medical treatment, accomplices, witnesses, etc.), victim/accused relationship, promptness of reporting, accused's prior record, methods of classifying complaints, subjective data (i.e., personal comments reported in files), and any information about the police processing of cases. Reports dropped at the police level were examined to determine whether the complainant or the police made the decision, and the reason for termination.
b) Crown Files Crown attorneys determine whether a complaint should be prosecuted, decide how the case should be prepared, and assess the likelihood of conviction. The Crown has the option of altering or reclassifying charges as they decide which charges to proceed with to obtain a conviction. They can also negotiate with defence counsel to secure a guilty plea from the accused in exchange for a reduction in the number or seriousness of the charges, or a recommendation for a lenient sentence.
Crown files provided details on charging, evidence, witnesses, plea-bargaining, preliminary hearings and preparation of complainants for court. Information regarding the termination of charges at this level was collected, including details on whether this decision was initiated by the Crown prior to a preliminary hearing, by a judge at a preliminary hearing, or by the complainant and why this decision was made. Crown files were also used to obtain information concerning the trial process, including the nature of final charges, disposition, sentencing and appeals. Seventy-eight cases in the pre-reform period and 230 cases in the post-reform period were tracked at the Crown level. Juveniles, who made up 17.6% of the 1981-1982 accused and 17.9% of the 1984-1985 accused, were not tracked beyond the police level as these cases are handled by Youth Court.
c) Court Files Most sexual assaults are handled as indictable offences, which allows the defendant to elect trial by Provincial Court Judge, Queen's Bench Judge or Queen's Bench Judge and Jury. The last two are favoured because the defendant is then entitled to a preliminary hearing. Often, the defendant agrees to plead guilty in exchange for a reduction in the number or seriousness of the charges, or for recommendation for a lenient sentence by the Crown. If the defendant pleads guilty prior to a trial, the plea is heard at a sentencing hearing. If a case proceeded to a sentencing appearance, preliminary hearing, and/or trial, it was tracked to the appropriate Court office (Provincial or Queen's Bench) and court documents were used to supplement information concerning the nature of final charges, dispositions, sentencing and appeals. Fifty-three cases at the pre-reform level and 178 at the post-reform level were tracked at the Court level.
Analysis of the filtering of cases through the Winnipeg justice system showed that the number of cases reported to the police increased by 66% following the implementation of C-127.(2) The conviction rate increased from 12.7% of the pre-reform cases to 29.1% of those post-reform cases that resulted in conviction (Gunn and Linden, 1993). While this increase seems to show that the legislation has had a significant impact on conviction rates, earlier research casts doubt on this conclusion. A study of the processing of sexual assault cases during 1976 and 1977 (Minch et al., 1987) showed a conviction rate of 28.9%, one almost identical to that found in the post-reform period. This would suggest that the 1981-1982 sample is atypical. This interpretation is supported by the fact that the only substantial difference between the two years is in the proportion of offenders not apprehended by the police. This was 27.9% of the pre-reform sample and 10.7% for the post-reform cases. In 1976-1977, the proportion not apprehended was 10.9%. Thus it does not appear that the revised sexual assault law has reduced the attrition in case processing in the courts, since the only difference occurred at the police level.
These data and the research cited earlier suggest that rape law reform has had a minimal impact on conviction rates. However, even if conviction rates have not increased, the legislation may have led to other changes in the fairness of the system by reducing the influence of legally irrelevant descriptions of the victim. The inclusion of clauses such as the rape shield law in the legislation makes it clear that one purpose of Bill C-127 was to reduce the impact of the victim's character on the outcome of a sexual assault case. The literature contains many examples of rape victims whose testimony was discredited by judges or juries who felt that they were responsible for their own victimization (Comack, 1992). Cases involving victims who dress inappropriately, who get picked up in bars, who hitchhike, who are prostitutes, who use drugs, or who engage in other behaviours departing from the norm of the stable, middle-class, married woman often result in acquittals (Gunn and Minch, 1988; LaFree, 1989).
There are two quite different reasons why the success of rape prosecutions often depends on these character issues. First, rape laws have historically been devised to protect male sexual property. The more valuable the property and the more powerful the property owner, the greater the penalty for rape. According to this patriarchal ideology, women only deserve protection if they fit the image of a dutiful wife or daughter raped by a stranger. Second, there are rarely witnesses in rape trials, so a judge or jury must decide the case on the basis of two conflicting versions of the event. If a defence lawyer can portray the defendant as a trustworthy, respectable person, and the victim as an immoral person whose story cannot be believed, a jury could have reasonable doubts about the guilt of the accused. This is particularly true when the defendant does not have to testify or be subjected to cross-examination, while the victim may be subjected to very tough questioning by the defence. The section of C-127 that restricts questioning concerning the past sexual history of the victim was intended to make this process more equitable.
Thus in addition to looking at the impact of Bill C-127 on conviction rates, we can also evaluate the legislation by comparing the processing of pre- and post-reform cases to determine whether extralegal variables involving the character of the victim made a difference in the founding, charging, prosecuting and convicting of sexual assault cases. This type of analysis has not been conducted in previous evaluations of the impact of rape reform legislation: it is important because even if conviction rates have not changed, the legislation may be considered at least partially successful if it results in the more equitable treatment by the courts of sexual assault cases.
Logistic Regression Analysis of Charges Laid and Convictions
In this section, we examine the effect of a number of variables on whether charges were laid at the police level and/or whether the accused was convicted. Since each of the two dependent variables takes on only two values, multivariate logistic regression is the appropriate statistical methodology. Logistic regression was run separately for the pre- and post-reform periods, thus allowing us to:
a) test whether each variable has a significant effect on charges being laid (and the accused being convicted), holding other variables constant, within the pre- and post-reform periods;
b) obtain numerical estimates of the increase (or decrease) in the probability of charges being laid (and of the accused being convicted) attributable to each of the explanatory variables, separately for the pre- and post-reform periods; and
c) compare the influence of the same variables across the two periods.
The explanatory variables used are described in the Appendix.
Logistic Regression on Charges Laid
Logistic regression was run on CHARGELD, defined as follows:
CHARGELD = i if a charge was laid at police level = 0 if a charge was not laid at police level
Table 1 presents the results of the logistic regression on CHARGELD, separately for the pre- and post-reform periods. The first and second columns present the logistic regression coefficients for the pre- and post-reform periods respectively. We note that OVKNOTH, NOOFVICT, BADCHAR and INJURIES were significant in explaining whether charges were laid in both the pre- and post-legislation periods. Furthermore, ATTMPTI and VAGELT18 were significant at the 5% level in the pre-legislation period only. The third and fourth columns present the estimated increase (decrease) in probability of charges being laid by the police, attributable to each significant variable for the pre- and post-legislation periods respectively. The interpretation of the third and fourth columns is as follows. Relative to the base case where all variables are equal to zero, that is, the accused is a stranger, the type of sexual contact is sexual intercourse or other (excluding attempted intercourse and touching/grabbing), the initial complaint was made by telephone to the police, there was only one complainant who was older than 17 years of age at the time of (first) assault, there was no evidence of bad character, there were no injuries, there were no witnesses, and the offence was reported by the complainant:
a) If the accused was known (other than a parent), the probability of charges being laid increases by 0.26 in the pre-legislation period and by 0.44 in the post-legislation period. This finding is partly due to the fact that when the accused is a stranger, he is often not apprehended and therefore no charges can be laid.
b) Relative to intercourse and other, if the sexual contact was attempted intercourse, the probability of charges being laid decreases by 0.15 in the pre-legislation period; however, this variable was not significant in the post-legislation period. This suggests that the removal of the penetration requirement may have led to more charges being laid in attempted rape cases.
c) If there is more than one complainant, the probability of charges being laid increases by 0.36 in the pre-legislation period and by 0.53 in the post-legislation period. The presence of more than one complainant provides a greater likelihood of identification of the accused and also provides corroboration of testimony.
d) Evidence of "bad character" decreases the probability of charges being laid by 0.21 in the pre-legislation period and by 0.24 in the post-legislation period. Thus, the legislation does not appear to have changed the importance of the complainant's character in the decision to charge. In many of these cases, however, it was found that the complainant refused to proceed, thus affecting charges being laid.
e) If the complainant received injuries as part of the assault, the probability of charges being laid increases by 0.17 in the pre-legislation period and by 0.15 in the post-legislation period. Injuries were somewhat less important in charging in the post-reform sample, which may be related to removal of corroboration requirement, but this difference was slight.
f) If the complainant was less than 18 years of age, the probability of charges being laid increased by 0.12 in the pre-legislation period;. this variable was not significant in the post-legislation period, however.
[TABULAR DATA FOR TABLE 1 OMITTED]
In the preliminary logistic regression tests, OALCOHOL was significant at the 5% level and the coefficient was positive, indicating that alcohol use by the offender increases the probability of charges being laid. Because of missing data, however, the inclusion of this variable in the equation reduced the sample upon which the logistic equation was estimated by more than half and it was therefore omitted in the regression equation to avoid the possibility of introducing sample selection bias in the equation.
Logistic Regression on Convictions
Logistic regression was run on CONVICTED, defined as follows:
CONVICTED = 1 if accused was convicted of sexual or non-sexual offence
= 0 if accused was not convicted of a sexual or nonsexual offence (i.e., charges not laid by the police, case stayed, found not guilty at trial)
Table 2 presents the results of the logistic regression on CONVICTED, run separately for the pre- and post-legislation periods. Juvenile offenders were excluded from this analysis as they were not handled by adult court. The first and second columns present the logistic regression coefficients for the pre- and post-legislation periods, respectively. We note that OVKNOTH, NOOFVICT, INJURIES, WITNESS and PARNTREP were significant in explaining whether the accused was convicted in the pre-legislation period. In the post-legislation period, OVKNOTH, INITOTHR, NOOFVICT and BADCHAR were significant in explaining whether the accused was convicted. The third and fourth columns present the estimated increase (decrease) in probability of an accused being convicted, attributable to each significant variable, for the pre- and post-legislation periods respectively. The interpretation of the third and fourth column is as follows. Relative to the base case where all variables are equal to zero, that is, the accused is a stranger, type of sexual contact is sexual intercourse or other (excluding attempted intercourse and touching/grabbing), the initial complaint was made by telephone to the police, there was only one complainant who was over 17 years of age at time of the (first) assault, there was no evidence of bad character, there were no injuries, there were no witnesses, and the offence was reported by complainant:
a) If the accused was known (other than parent), the probability of the accused being convicted increases by 0.03 in the pre-legislation period and by 0.22 in the post-legislation period. This finding is related to the fact that when the accused is a stranger, he is often not apprehended and therefore cannot be convicted.
b) Relative to the case where the initial complaint was telephoned to the police, other methods of reporting the initial complaint (excluding personal reporting at the police station or at the scene of the offence) decreased the probability of a conviction by 0.21 in the post-legislation period; this variable was not significant in the pre-legislation period.
c) If there is more than one complainant, the probability of a conviction increases by 0.03 in the pre-legislation period and by 0.38 in the post-legislation period.
d) Evidence of"bad character" decreases the probability of a conviction by 0.29 in the post-legislation period; this variable was not significant in the pre-legislation period. This may not indicate a bias of the post-legislation courts, as many complainants withdrew from the proceedings. For example, 14 out of 20 prostitutes in the post-reform sample refused to cooperate or did not show up in court.
e) If the complainant suffered injuries as a result of the assault, the probability of a conviction increased by 0.04 in the pre-legislation period; this variable was not significant, however, in the post-legislation period. This finding offers some very modest evidence that the removal of corroboration may have an effect on conviction, as does f).
f) The presence of a witness increased the probability of a conviction by 0.02 in the pre-legislation period; this variable was not significant in the post-legislation period.
g) If the offence was reported by a parent (relative to the complainant reporting the offence), the probability of a conviction increases by 0.03 in the pre-legislation period; this variable was not significant in the post-legislation period.
[TABULAR DATA FOR TABLE 2 OMITTED]
It appears that different variables were affecting convictions in the pre- and post-legislation periods. The results obtained in the pre-legislation period must be examined with caution, however, as the overall proportion of convictions was low and significance tends to be based on a few cases.
In the preliminary logistic regression runs, PHYSICAL was significant at the 5% level for the pre-legislation period and the coefficient was positive, indicating that the use of physical force by the defendant increased the probability of a conviction. Including this variable in the equation reduced the sample upon which the logistic equation was estimated by more than half (reducing the number of convictions to 20) and it was therefore omitted in the regression equation presented here.
In the logistic regression run shown in Table 2, the variables affecting court decisions are not isolated from those affecting the police because all cases were entered into the regression, including those cases in which charges were not laid. To isolate the effect of each variable on conviction alone, the logistic regressions were rerun, excluding all the cases where charges were not laid. Table 3 presents the results of the logistic regression on CONVICTED, run separately for the pre- and post-legislation periods. Juvenile offenders were also excluded from this analysis as those cases were not handled by adult court. The first and second columns present the logistic regression coefficients for the pre- and post-legislation periods respectively. We note that only INJURIES was significant in explaining whether the accused was convicted in the pre-legislation period. In the post-legislation period, NOOFVICT and BADCHAR were significant in explaining whether the accused was convicted. The third and fourth columns present the estimated increase (decrease) in probability of the accused being convicted, attributable to each significant variable, for the pre- and post-legislation periods respectively. The interpretation of the third and fourth column is as follows. Relative to the base case where all variables are equal to zero, that is, the accused is a stranger, the type of sexual contact is sexual intercourse or other (excluding attempted intercourse and touching/grabbing), the initial complaint was made by telephone to the police, there was only one complainant, the complainant was more than 18 years of age at time of the (first) assault, there was no evidence of bad character, there were no injuries, there were no witnesses, and the offence was reported by the complainant:
a) If there is more than one complainant, the probability of a conviction increases by 0.22 in the post-legislation period; this variable was not significant in explaining conviction in the pre-legislation period.
b) Evidence of bad character decreases the probability of a conviction by 0.27 in the post-legislation period; this variable was not significant in the pre-legislation period.
c) If the complainant suffered injuries as a result of the assault, the probability of a conviction increased by 0.18 in the pre-legislation period; this variable was not significant in the post-legislation period.
[TABULAR DATA FOR TABLE 3 OMITTED]
It should be noted that the results in the pre-legislation period must be interpreted with caution as the overall proportion of convictions is low (N = 38) and significance tends to be based on a few cases.
The results presented here show that the impact of Canada's rape reform legislation has been modest. The conviction rate does not appear to have increased, and variables reflecting the character of the victim still have a significant impact on the laying of charges and on the likelihood of conviction.
Why did the legal change have such little impact? A major reason is that changing the law does not necessarily mean that the criminal justice system can or will quickly alter the way in which it operates. Boyle points out that "one of the key functions of law reform is the exertion of control over the decision makers at various levels of the criminal justice system" (1984: 127). However, these functionaries may resist change; the decision-making context remains the same and discretion is still in the hands of the same actors who were responsible for the criminal justice system before the legislation was enacted. Boyle also notes that the legal change was aimed at the trial stage, and that it may have had little impact on victim reporting, police and prosecutorial discretion, or sentencing. Since very few sexual assault cases actually go to trial, the impact of the legislation was not likely to be great.
The success of rape law reform may also have been limited by the fact that sexual assault cases are inherently difficult to prosecute, as the crime typically takes place in private and there are rarely witnesses other than the victim and the accused. Since the prosecution must establish proof beyond a reasonable doubt when evidence may consist simply of one person's word against another's, factors such as corroboration and recency of complaint are important in presenting a strong case even if they are not legally required.
In looking at the low rates of conviction for sexual assault, we should keep in mind that the criminal justice system does not efficiently prosecute criminals. Too often, critics look at the low conviction rate for sexual assault and assume that this offence is atypical. However, conviction rates are low for almost all offences. based on their comparison of the processing of rape cases with homicide, robbery, assault and burglary in several California counties, Galvin and Polk concluded that while there is a high rate of attrition for rape cases, it is not unique, as attrition is "a distinctive feature of American justice across offences" (1983: 152). In Canada, Renner and Sahjpaul (1986) found that the percentage of reported cases resulting in charges was almost identical for physical assault and sexual assault. The percentage of cases cleared that were also charged, however, was considerably higher for sexual assault (64%) than for physical assault (42%).
The situation was not, however, entirely negative. As we have reported elsewhere, the legislation did have a positive impact on attitudes and practices toward sexual assault victims (Gunn and Linden, 1988). More complainants in the post-reform period were reporting cases that differed from the stereotype of violent attacks by strangers. For example, more prostitutes reported being victimized (though 70% of them later refused to cooperate or did not show up in court) and a smaller proportion of cases involved offenders who were strangers to the victims. Physical force was reported less often in the post-reform sample. Interviews with criminal justice personnel and victims showed positive attitudes toward Bill C-127. Police, prosecutors, judges and sexual assault counsellors all felt the legislation was an improvement, and victims generally felt they had been treated fairly by the system.
Respondents agreed that the courts were following the new law. This perception was supported by our court observation: victims were not questioned about their previous sexual behaviour, and when the defence attempted to pursue this line of questioning, the Crown and judge intervened.
Despite the limited impact the legislation had on the processing of sexual assault cases, we would argue that its passage was still important. The law recognizes the complaints that had been raised about the treatment of rape victims. The feminist critique that the law served as a mechanism for controlling the behaviour of women, particularly those whose conduct did not conform to middle-class standards, was a legitimate one. The previous practice was degrading and unresponsive to victims and was clearly unfair. The law assumed that women were likely to lie about being raped and that special legal restrictions were needed to ensure that accused rapists received a fair trial. Although the statistical impact shown here was small, legislative change was a necessary step in shifting the focus in sexual assault cases from the reputation and behaviour of the victim to the criminal actions of the offender.
The passage of Bill C-127 is an affirmation of the rights of women in the criminal justice system, but legislation cannot solve the problem of violence against women. The only real solution lies in changing communities to ensure they are just, fair and safe for all citizens (Macleod, 1989; 1994). However, Bill C-127 does provide a formal recognition of social change, which can contribute to the long-term shift in attitudes that is necessary in moving toward a safer society for all.
Definition of Variables
OVPRNT = 1 If the accused (offender-victim relationship) was a parent (father, mother, adoptive parent, step/common-law parent or foster parent)
= 0 Otherwise
OVKNOTH = 1 If complainant (victim) had knowledge of the accused but the accused was not a parent
= 0 Otherwise
TOUCHGR = 1 If type of sexual contact involved touching/grabbing
= 0 Otherwise
ATTMPTI = 1 If type of sexual contact involved attempted intercourse
= 0 Otherwise
INITPERS = 1 If initial complaint was made personally at the police station or at the scene of the offence
= 0 Otherwise
INITOTHR = 1 If initial complaint was not made personally at the police station, at the scene of the offence or by telephone to the police
= 0 Otherwise
NOOFVICT = 1 If there was more than one complainant (victim)
= 0 If there was only one complainant (victim)
BADCHAR[*] = 1 If there was evidence alleging that the complainant was of "bad character" or "promiscuous"
= 0 If there was no evidence alleging "bad character"
INJURIES = 1 If there were injuries suffered by the complainant
= 0 If there were no injuries suffered by the complainant
WITNESS = 1 If there was a witness to the incident
= 0 If there was no witness to the incident
PARNTREP = 1 If offence was reported by father/surrogate or mother/surrogate
= 0 Otherwise
OTHERREP = 1 If offence was reported by someone other than mother, father or complainant (i.e., friend, boyfriend, other relative, neighbour, school, hospital, social worker, spouse)
= 0 Otherwise
VAGELT18[**] = 1 If the complainant (victim) was less than 18 years of age at time of (first) assault
= 0 If complainant (victim) was more than 17 years of age
PHYSICAL = 1 If physical force was used by the accused
= 0 If no physical force was used
VICRES = 1 If complainant (victim) resisted accused
= 0 If complainant (victim) did not resist
VALCOHOL = 1 If alcohol was present in complainant (victim)
= 0 If alcohol was not present in complainant (victim)
OALCOHOL = 1 If alcohol was present in accused
= 0 If alcohol was not present in accused
* The inclusion of this variable is not intended as a moral judgment on the part of the researchers, but is meant to incorporate judgmental criteria that have been used by decision makers in the criminal justice system.
** The last four variables were missing in a large portion of the cases, thus reducing the sample size considerably. Although they were included in preliminary tests, they are not reported here, as the severely reduced sample may bias the other estimates. However, if they were significant in the preliminary runs, this is indicated in the text.
This is a revised version of a paper presented at the annual meeting of the American Society of Criminology. Funding for this research was provided by the Department of Justice and by the Solicitor General's Contributions Grant to the Criminology Research Centre, University of Manitoba. The authors would like to thank Don Sabourin for his assistance and the reviewers for their very helpful comments. The manuscript of this article was submitted in April 1994 and accepted in August 1995.
1. Bill C-127 was a response to many years of lobbying by women's organizations and to the 1982 enactment of the Charter of Rights and Freedoms. Los (1994) provides a fascinating account of the development of this legislation.
2. This is consistent with the findings of Renner and Sahjpaul and those of Roberts. Roberts and Gebotys (1993) have shown t.hat this increase is likely due to the impact on public attitudes of the passage of the 1983 legislation and the publicity surrounding the new law. They tested a variety of other hypotheses, including the possibility of changes in police recording patterns, and found that none could have accounted for a change of this magnitude.
Boyle, C.L.M. 1984. Sexual Assault. Toronto: The Carswell Company.
Clark, L. and D. Lewis. 1977. Rape: The Price of Coercive Sexuality. Toronto: The Women's Press.
Clark, S. and D. Hepworth. 1994. "Effects of reform legislation on the processing of sexual assault cases." In Confronting Sexual Assault:A Decade of Legal and Social Change. J.V. Roberts and R.M. Mohr (eds.). Toronto: University of Toronto Pr., pp. 113-35.
Comack, E. 1992. "Women and crime." In Criminology: A Canadian Perspective. 2d ed. R. Linden (ed.). Toronto: Harcourt, Brace, Jovanovich, pp. 127-62.
DeKeseredy, W. and R. Hinch. 1991. Women Abuse: Sociological Perspectives. Toronto: Thompson Educational Publishing.
Department of Justice. 1990. Sexual Assault Legislation in Canada: An Evaluation - Overview. Ottawa: Department of Justice.
Galvin, J. and K. Polk. 1983. "Attrition in case processing: Is rape unique?" Journal of Research in Crime and Delinquency, Vol. 20, pp. 126-53.
Gunn, R. and R. Linden. 1988. Evaluation of Bill C-127: Synthesis Report. Winnipeg: Prairie Research Associates.
Gunn, R. and R. Linden. 1993. "Evaluating the impact of legal reform: Canada's new sexual assault laws." In Evaluating Justice: Canadian Policies and Programs. J. Hudson and J. Roberts (eds.). Toronto: Thompson Educational Publishing, pp. 135-54.
Gunn, R. and C. Minch. 1988. Sexual Assault: The Dilemma of Disclosure, the Question of Conviction. Winnipeg: University of Manitoba Pr.
Hinch, R. 1996. "Sexual violence and social control." In Social Control in Canada. B. Schissel and L. Mahood (eds.). Toronto: Oxford Univ. Pr.
LaFree, G.D. 1989. Rape and Criminal Justice: The Social Construction of Sexual Assault. Belmont, Calif.: Wadsworth.
LeBeau, J.L. 1987. "Statute revision and the reporting of rape: An associational or causal relationship." Paper presented at the Annual Meeting of the American Society of Criminology, Montreal.
Loh, W.D. 1980. "The impact of common law and reform rape statutes on prosecution." Washington Law Review, Vol. 55, pp. 543-625.
Los, M. 1994. "The struggle to redefine rape in the early 1980's." In Confronting Sexual Assault: A Decade of Legal and Social Change. J.V. Roberts and R.M. Mohr (eds.). Toronto: University of Toronto Pr., pp. 20-56.
MacLeod, L. 1989. The City for Women: No Safe Place. Ottawa: Secretary of State.
MacLeod, L. 1994. Understanding and Charting our Progress Toward the Prevention of Women Abuse. Ottawa: Family Violence Prevention Division, Health Canada.
Marsh, J.C., A. Geist and N. Caplan. 1982. Rape and the Limits of Law Reform. Boston: Auburn House.
Minch, C., R. Linden and S. Johnson. 1987. "Attrition in the processing of rape cases." Canadian Journal of Criminology, Vol. 29, pp. 389-404.
Polk, K. 1985. "Rape reform and criminal justice processing." Crime and Delinquency, Vol. 31, pp. 191-205.
Renner, K.E. and S. Sahjpaul. 1986. "The new sexual assault law: What has been its effect?" Canadian Journal of Criminology, Vol. 28, pp. 407-13.
Roberts, J.V. 1990. Sexual Assault Legislation in Canada: An Analysis of National Statistics. Ottawa: Department of Justice.
Roberts, J.V. and R.J. Gebotys. 1993. "Evaluating criminal justice legislation: Effects of rape reform in Canada." In Evaluating Justice: Canadian Policies and Programs. J. Hudson and J. Roberts (eds.). Toronto: Thompson Educational Publishing, pp. 155-69.
Spohn, C. and J. Horney. 1992. Rape Law Reform:A Grassroots Revolution and Its Impact. New York: Plenum Press.…
Questia, a part of Gale, Cengage Learning. www.questia.com
Publication information: Article title: The Impact of Law Reform on the Processing of Sexual Assault Cases. Contributors: Gunn, Rita - Author, Linden, Rick - Author. Journal title: The Canadian Review of Sociology and Anthropology. Volume: 34. Issue: 2 Publication date: May 1997. Page number: 155+. © 1999 Canadian Sociology and Anthropology Assn. COPYRIGHT 1997 Gale Group.
This material is protected by copyright and, with the exception of fair use, may not be further copied, distributed or transmitted in any form or by any means.