|Title:||The impact of rape law reform on the processing of simple and aggravated rape cases.|
|Abstract:||Focuses on the impact of rape law reform on the processing of rape cases in the United States. Background on the study; Rape law reform in Michigan; Research design; Sampling and data collection procedures; Selection of cases for analysis; Simple versus aggravated rape cases; Case outcomes.|
|Full Text Word Count:||9043|
|Database:||Academic Search Premier|
There has been significant change in the laws and the rules of evidence relevant to the crime of rape during the past two decades. Reformers criticized traditional rape laws that defined rape narrowly; these laws would require both proof that the victim resisted her attacker and corroboration of the victim's testimony. The defense could use evidence of the victim's past sexual behavior to prove that she consented, or to impeach her credibility. Critics charged that these laws shifted the focus of a rape case from the behavior of the offender to the character and behavior of the victim. They argued that the laws discouraged rape victims from reporting the crime to the police and erected significant barriers to the successful prosecution of rape cases.
In response to these concerns, states enacted a variety of reforms. Some states replaced the single crime of rape with a series of gender-neutral offenses defined by the presence or absence of aggravating circumstances such as use of a weapon, injury to the victim, or commission of a contemporaneous felony. Many states eliminated legal requirements that the victim physically resist her attacker and that her testimony be corroborated. Most states also enacted rape shield laws restricting the use of evidence of the victim's reputation or past sexual behavior.
Reformers predicted that these legal changes would produce a number of instrumental effects. They expected the new laws to improve the treatment of rape victims and thus to prompt more rape victims to report the crime to the police. They also predicted that broadening the definition of rape and changing the rules of evidence applied in rape cases would make arrest, prosecution, conviction, and incarceration for rape more likely.
Studies evaluating the impact of the rape law reforms suggest that reformers' expectations were overly optimistic. Most of these studies have not found the anticipated effects on reports of rape or the processing of rape cases, although two studies found that the reforms enacted in Michigan led to increases in the number of arrests and convictions for rape. Studies conducted in King County (Seattle), Washington, California, and Nebraska found that the legal changes did not produce the expected results. Our study of the impact of rape law reform in six jurisdictions similarly found minimal effects. We used time-series data to evaluate the effect of the reforms on reports of rape and the processing of rape cases in Detroit, Chicago, Philadelphia, Atlanta, Houston, and Washington, D.C. We found that the reforms produced significant effects only in Detroit, and even there the impact was limited. The strong and comprehensive reforms enacted in Michigan resulted in an increase in the number of reports of rape and in the percentage of cases bound over for trial, but had no effect on the conviction rate or sentencing.
The results of these studies suggest that changing the definition of rape and the rules of evidence applied in rape cases has not produced the dramatic results anticipated by reformers. It is possible, however, that the reforms have produced more subtle results. The studies conducted to date have evaluated the reforms' impact on the processing of rape cases generally. These studies have not explored the possibility that the reforms have produced changes in the types of rape cases entering the criminal justice system; they also have not explored the possibility that the reforms have had differential effects on rape cases of varying degrees of seriousness.
Our earlier research on rape law reform in Detroit provides evidence in support of a more subtle impact. In interviews, criminal justice officials suggested that the rape law reforms had increased the likelihood of prosecution and conviction in rape cases in which the victim and the defendant were acquainted; more generally, they suggested that the odds of successful prosecution of "less serious" or "borderline" rape cases had improved in the post-reform period.
These speculations were partially confirmed in a follow-up study examining the characteristics of rape cases bound over for trial in Detroit from 1973 through 1984. We compared pre- and post-reform cases and found that a larger proportion of the post-reform cases involved questions about the victim's moral character, evidence of risk-taking behavior by the victim, and a victim and offender who were acquainted. We concluded that these changes reflected changes in the types of cases reported to the police, as well as changes in the screening criteria used by police and prosecutors.
In this Article we continue our search for more subtle impact by comparing the types of cases bound over for trial in Detroit Recorder's Court in the pre-reform and post-reform periods and by examining case outcomes for different types of rape cases in the two time periods. In doing so, we rely on the distinction made by Kalven and Zeisel and by Estrich between "simple" and "aggravated" rapes. Aggravated rapes are incidents involving extrinsic violence, multiple assailants, or no prior relationship between victim and offender; simple rapes are incidents with none of these aggravating circumstances.
Both Kalven and Zeisel and Estrich assert that aggravated and simple rape cases are treated differently by the criminal justice system. Kalven and Zeisel, for example, found jury conviction rates nearly four times higher in aggravated than in simple rape cases. They also found that judges were much more likely to disagree with jury verdicts in simple rape cases, believing that juries in these types of cases relied too heavily on victim characteristics and often acquitted in spite of sufficient evidence to convict.
Estrich similarly maintains that the handling of rape cases depends upon the type of rape; simple rape cases are less likely to be reported to the police, less likely to be prosecuted, and less likely to result in a conviction. She argues, in fact, that historically the processing of rape cases has not been characterized by indiscriminate sexism, but that there has been and is:
Estrich also suggests that traditional rape law provisions represented "a set of clear presumptions applied against the woman who complains of simple rape." She argues that because the essential features of aggravated rape cases--an attack by a stranger, multiple assailants, the use of a weapon, or injury to the victim--met the requirements of "real rape," there was no reason to distrust the victim in these cases. If the rape was a "real rape," in other words, it was less likely that the prosecutor or judge would demand proof that the victim physically resisted her attacker, require corroboration of her testimony, or allow the defense to introduce evidence of her past sexual behavior.
If Estrich is correct, then it follows that most of the rape law reforms have been directed at simple rape cases, and thus the greatest impact should have been felt in these cases. In our previous six city study, we assessed the reforms' impact only in terms of overall numbers of reports and rates of prosecution, conviction, and incarceration; we were not able to compare case outcomes for different types of rape cases. More to the point, we were not able to compare case outcomes for aggravated and simple rape cases in the pre- and post-re-form periods.
In this Article, we use data on the characteristics of rape cases bound over for trial in Detroit Recorder's Court to test the hypothesis that the rape law reforms have had a significant effect on the processing of simple (but not aggravated) rapes. More specifically, we test the hypothesis that the proportion of simple rape cases bound over for trial will be greater in the post-reform period than in the pre-re-form period. We also test the hypothesis that simple rape cases will be taken more seriously in the post-reform period. We hypothesize that simple rape cases will be less likely to be dismissed, more likely to re-suit in a conviction, and more likely to result in a severe sentence in the post-reform period than in the pre-reform period. Because we believe that aggravated rapes have always been taken seriously, we do not expect to find similar changes for aggravated rape cases.
We suggest that the rape law reforms have blurred the distinctions between simple and aggravated rape cases. We argue that simple rape cases were treated differently than aggravated rape cases in the pre-reform period, but that these differences have diminished in the post-reform period. Accordingly, we hypothesize that the "degree of aggravation" will have less impact on the outcome of rape cases in the post-reform period than in the pre-reform period.
We focus on the impact of reforms enacted in Michigan. Of the six jurisdictions included in our earlier study, Michigan adopted the strongest and most comprehensive reforms. In fact, the comprehensive Michigan statute implemented in 1975 is regarded by many as a model rape reform law.
The Michigan statute redefines rape and other forms of sexual assault by establishing four degrees of gender neutral criminal sexual conduct based on the seriousness of the offense, the amount of force or coercion used, the degree of injury inflicted, and the age and incapacitation of the victim. The statute extends the reach of the sexual assault laws to acts (sexual penetration with an object) and persons (men and married persons who are legally separated) not covered by the old laws.
The Michigan law also delineates the circumstances that constitute coercion, lists the situations in which no showing of force is required (for example, when the victim is physically helpless or mentally defective), and states explicitly that the victim need not resist the accused. Because evidence of coercion is seen as tantamount to non-consent, the law effectively eliminates the requirement that the prosecutor prove that the victim did not consent; the burden of proving the victim acquiesced to the act falls to the defendant. The Michigan law further modifies the rules of evidence by stating that corroboration of the victim's testimony is not required.
The 1975 Michigan reform also included a strong rape shield law that generally prohibits the introduction of evidence of the victim's past sexual conduct. The prohibition applies to evidence of specific instances of sexual activity, reputation evidence, and opinion evidence. There are only two exceptions to the shield: evidence of the victim's past sexual conduct with the defendant can be admitted, but only if it is material to a fact at issue (generally consent) and if its inflammatory or prejudicial nature does not outweigh its probative value; evidence of prior sexual activity with persons other than the defendant is admissible only to show the source of semen, pregnancy, or disease.
It is clear that the strong and comprehensive reform law enacted in Michigan was more than a symbolic response to a vocal constituency clamoring for change. The law defined new crimes, mandated new procedures, and limited the discretion of criminal justice officials. It also was accomplished in one major revision of state codes. Reformers clearly expected these changes to influence the processing and disposition of rape cases? We address this expectation below.
For our earlier study of the impact of rape law reform we collected court records data on all rape cases (N=24,000) bound over for trial from 1970 through 1984 in six jurisdictions. For this project we obtained detailed information on case characteristics for a random sample of cases in Detroit. We selected Detroit because we wanted to understand the mechanisms through which the statutory changes produced their impact and Detroit was the only jurisdiction in our original study where the reforms had significant effects. We also selected Detroit because our interviews with criminal justice officials there led us to believe that the strong and comprehensive reforms adopted in Michigan had subtle effects on case processing that could not be detected with the data obtained for our first project.
We recognize that focusing on Detroit, while allowing us to test for effects that might have been masked in our earlier study, biases our research design. As noted above, the reforms enacted in Michigan were unusually strong and comprehensive. If we find that these reforms produced significant effects, we obviously cannot then conclude that all rape law reforms will produce similar effects. On the other hand, if we discover that the strong reforms enacted in Michigan did not produce the predicted impact on simple rape cases, we might conclude that weaker reforms would be even less likely to generate this result.
A. SAMPLING AND DATA COLLECTION PROCEDURES
We selected a random sample of all sexual assaults bound over for trial in Detroit Recorder's Court from 1970 through 1984. Because we intended to compare the outcome of cases involving strangers and acquaintances, and because cases involving acquaintances are likely to result in a less serious charge than "rape," we selected the sample from defendants charged with a variety of offenses. We included defendants charged with rape, sodomy, gross indecency, or assault with intent to rape in the pre-reform period and defendants charged with first, second, or third degree criminal sexual conduct in the post-reform period. There were no fourth degree criminal sexual conduct cases (CSC4) in Recorder's Court since CSC4 is a high misdemeanor under Michigan law.
Because we intended to compare the outcome of cases in the pre-reform and post-reform periods, we stratified the sample to ensure an adequate number of cases in each time period. There were 835 sexual assault cases in the pre-reform period and 3,869 cases in the post-reform period. The sample included all of the cases in the pre-reform period (N=835) and 25% of the cases filed each year in the post-re-form period (N=968).
Detailed information on the cases included in the sample was obtained from case files maintained by the Sex Crimes Unit of the Detroit Police Department. Data. collectors read all documents and reports in the case file and recorded information about the case, the complainant, the offender, and the incident on an op-scan form designed for the project. This information was then merged with the case outcome data collected for the earlier study.
Unfortunately, data collectors were not able to locate police case files for all of the cases included in the sample. Most of the files for sexual assaults reported in 1970, 1971, and 1972 had been destroyed, and about 25% of the files for more recent cases also were missing. Data collectors were able to obtain information on 361 pre-reform cases and 815 post-reform cases.
B. SELECTION OF CASES FOR ANALYSIS
To compare the outcome of simple and aggravated rape cases before and after the reforms were enacted, we need to ensure that the crimes included in the two periods are equivalent. This is problematic, given that the 1975 reform included significant changes in the definition of the crime of rape. As noted above, the Michigan criminal sexual conduct statute is more inclusive than the previous statute; it covers acts (e.g., oral and anal sex, penetration with an object, fondling, incest) and persons (men and legally separated spouses) not covered by the old rape law.
To ensure the equivalence of crimes in the two time periods, we included all pre-reform cases where the most serious charge was rape, sodomy, or gross indecency. We included sodomy and gross indecency in addition to rape because these charges define sexual acts that are incorporated in first and third-degree criminal sexual conduct under the new laws. For the post-reform period, we included all cases where the most serious charge was first-degree or third-degree criminal sexual conduct and the most serious charge under the old laws would have been either rape, sodomy, or gross indecency. We excluded second-degree criminal sexual conduct (sexual contact without penetration) from the post-reform sample because this was a crime that would not have been defined as rape before the law changed. Use of these criteria resulted in the selection of 812 cases; this includes 279 pre-reform cases and 533 post-reform cases.
C. SIMPLE VERSUS AGGRAVATED RAPE CASES
This Article focuses on the impact of rape law reform on the processing of simple and aggravated rape cases. In defining these types of cases, we rely on the distinctions made by Kalven and Zeisel and Estrich. Aggravated rapes are cases in which the victim and the offender were strangers, the offender used a weapon, the offender injured the victim, or there were multiple assailants. Simple rape cases are cases with none of these aggravating circumstances.
Although both Kalven and Zeisel and Estrich categorized rape cases as either aggravated or simple rapes, we argue that there are degrees of aggravation. A rape case in which a stranger held a gun to the victim's head or a knife to her throat is qualitatively different from a case in which the victim was attacked by an unarmed stranger. A case with all four aggravating circumstances is more serious than a case with only one of the four aggravating conditions.
We created a variable measuring the degree of aggravation by summing the four aggravating circumstances (use of a gun or knife, injury to the victim, multiple assailants, victim and offender strangers). The resulting "scale of aggravation" ranges from 0 (no aggravating factors) to 3 (three or four aggravating factors). There are 179 (22.0%) cases coded 0, 225 (27.7%) coded 1,268 (33.0%) coded 2, and 140 (17.2%) coded 3.
D. CASE OUTCOMES
We examine five case processing outcomes. We first analyze the decision to dismiss all of the charges against the defendant. Since all of the cases in the sample are cases that were bound over for trial, this reflects decisions made after arraignment. Either the defense attorney or the prosecutor can submit a motion to dismiss prior to trial or a motion to adjourn at the trial; the judge hearing the case will grant or deny the motion. This variable is coded 1 if all charges were dismissed and 0 if all charges were not dismissed.
We also analyze two measures of the likelihood of conviction. The first variable focuses on conviction at trial. This variable is coded 1 if the defendant was convicted at trial and 0 if the defendant was acquitted at trial; defendants who had all charges dismissed or who pied guilty are excluded from the analysis. The second variable measures whether the case resulted in a conviction (by plea or at trial). This variable is coded 1 if the defendant pled guilty or was found guilty at trial and is coded 0 if the charges were dismissed or the defendant was acquitted.
Our fourth dependent variable is a dichotomous variable indicating whether a convicted defendant was incarcerated. The final case outcome is the sentence (in months) imposed on defendants who were incarcerated. These variables, their codes, and their frequencies are displayed in Table 1.
E. INDEPENDENT VARIABLES
In testing our hypothesis that the degree of aggravation associated with a rape case will have a greater impact on case outcomes in the pre-reform period than in the post-reform period, we controlled for variables that have been shown to affect the processing of criminal cases generally and of sexual assault cases specifically. These variables also are displayed in Table 1.
We included a number of variables measuring the seriousness of the offense. We used the scale of aggravation described above to create four dummy variables: Aggrp0 (no aggravating factors), Aggrp1 (one aggravating factor), Aggrp2 (two aggravating factors), and Aggrp3 (three or four aggravating factors). In all of the analyses, Aggrp3 is the omitted category.
Other factors relating to the seriousness of the offense are the number of charges against the defendant, the most serious conviction charge, whether the crime involved vaginal-penile penetration, and whether the offender used threats or force to subdue the victim.
We included two variables measuring the strength of evidence in the case. We reasoned that the odds of successful prosecution would improve if there was a witness to the incident or physical evidence (semen, fingerprints, blood stains, hair or skin samples) that could be used at trial. In analyzing the two sentencing decisions, we controlled for whether the defendant pled guilty or not. In all of the analyses, we included a variable that indicates whether the case was decided in the pre-reform period or in the post-reform period. Cases bound over for trial from January of 1970 through March of 1975 were classified as pre-reform cases; cases bound over from April of 1975 through December of 1984 were classified as post-reform.
We included six victim characteristics that have been shown to affect the processing of sexual assault cases. Previous research has demonstrated that sexual assault case outcomes are affected by the victim's socio-economic status, character, and relationship with the accused. Studies have shown that the treatment of men accused of sexual assault is influenced by victim "misconduct" such as hitchhiking or drinking, by the victim's reputation, and by the victim's age, occupation, and education.
We controlled for the victim's race and age. We also included a measure of the victim's "risk-taking" behavior; this variable was coded 1 if the police file indicated that at the time of the incident the victim was walking alone late at night, was hitch-hiking, agreed to accompany the offender to his residence, invited the offender to her residence, was in a bar alone, was using alcohol, or was using drugs. In addition, we controlled for whether the victim screamed during the attack, whether the victim physically resisted the offender, and whether the victim reported the crime to the police within one hour.
We also included controls for three offender characteristics--race, age, and prior felony convictions--that have been shown to affect processing decisions in criminal cases.
F. ANALYTIC PROCEDURES
Because there were not enough cases for a time-series analysis, we used a before-and-after design to test our hypotheses. We used a two-stage analytic procedure to test our hypothesis that the rape law reforms have had a significant effect on simple (but not aggravated) rape cases. We first computed chi-square statistics to compare the proportion of simple and aggravated cases in the pre-reform and post-reform periods and to compare case outcomes for each of the four types of cases in the two time periods. We then used either logistic regression or OLS regression to analyze the impact of the degree of aggravation on case outcomes in the two time periods; logistic regression was used to analyze the four dichotomous variables and OLS regression was used to analyze sentence length, an interval-level variable.
As noted above, we are hypothesizing that the degree of aggravation characterizing a rape case will have less impact on the outcome of the case in the post-reform period than in the pre-reform period. To test this hypothesis, we created interaction terms by multiplying the pre-reform/post-reform variable by the three dummy variables measuring the degree of aggravation in the case. For each outcome measure we estimated a regression equation in which all of the independent variables were first entered in a block, followed by a block containing the three interaction terms. We looked at the chi-square (at the R[sup 2] change in the OLS analysis) for improvement in the model after adding the interaction terms to test the null hypothesis that the addition of the interaction terms would not produce significant improvement in the goodness of fit of the model (p < .05).
If adding the interaction terms did not produce significant improvement in the fit of the model, we concluded that the effect of degree of aggravation did not differ in the pre- and post-reform periods. In that case, we present the regression coefficients obtained from the additive model. If adding the interaction terms did significantly improve the fit of the model, we concluded that the degree of aggravation had different effects in the two time periods, and we present the results obtained from the interactive model.
A. SIMPLE RAPES IN THE PRE-REFORM AND POST-REFORM PERIODS
We hypothesized that the proportion of simple rape cases bound over for trial would be greater in the post-reform period than in the pre-reform period. This hypothesis was confirmed. As shown in Figure 1, the percentage of simple rape cases increased from 17.6% to 24.4%. There was a similar decrease in the proportion of cases with two aggravating circumstances (from 38.4% to 30.0%). The proportion of cases in the other two categories remained fairly stable.
These results suggest that more cases of rape by unarmed acquaintances are "getting into the system" in the post-reform period.
Because all of the cases in our sample are cases that were bound over for trial, we do not know if this reflects changes in the characteristics of offenses, changes in the types of cases reported to the police, or changes in the criteria used by police and prosecutors in screening cases. We address these issues below.
B. CASE OUTCOMES IN THE PRE-REFORM AND POST-REFORM PERIODS
Our second hypothesis is that simple rape cases would be taken more seriously in the post-reform period. We predicted that simple rapes would be less likely to be dismissed, more likely to result in conviction, and more likely to result in a severe sentence in the post-re-form period than in the pre-reform period.
Case outcomes for the four types of rape cases in the two time periods are presented in Table 2. These results do not support our hypothesis. Although there were changes in case outcomes for simple rapes, the differences, with one exception, were not statistically significant. Moreover, the change in the rate of conviction at trial (from 16.7% to 50.0%), which was statistically significant, must be interpreted with caution; there were only twelve cases of simple rape that went to trial in the pre-reform period.
C. CASE OUTCOMES AND DEGREE OF AGGRAVATION; MULTIVARIATE ANALYSES
Our final hypothesis focuses on differences in the treatment of simple and aggravated rape cases in the pre-reform and post-reform periods. We suggested that one consequence of the rape law reforms has been a blurring of the distinctions between simple and aggravated rape; we argued that under the new laws, simple rapes are more likely to be viewed as "real rapes." We therefore hypothesized that the degree of aggravation would have less impact on the outcome of rape cases in the post-reform period than in the pre-reform period.
Because of the small number of simple rapes that went to trial in the pre-reform period, we were not able to model this decision. The results of our analyses of the other four case outcomes are presented in Tables 3 through 6. At the bottom of each table we present the chi-square value reflecting the improvement of the interactive model over the additive model. We found that adding the aggrp x reform interaction terms did not result in significant improvement in the goodness of fit of the model for any of the four case outcomes. For each of these decisions, the degree of aggravation had similar effects before and after the enactment of rape reform legislation.
The results of our analysis of the decision to dismiss the charges against the defendant are presented in Table 3. We found not only that the degree of aggravation did not have different effects in the two time periods, but that the degree of aggravation had no effect whatsoever on the likelihood of charge: dismissal. Charge dismissal was less likely if the victim was black, if the victim screamed during the attack, and if the victim reported the crime within one hour. The likelihood of charge dismissal was greater if the defendant was black; it was smaller if the defendant was charged with more than one crime.
As shown in Table 4, the number of aggravating circumstances did not have differential effects on the overall conviction rate (by plea or by trial) in the two time periods. There was a greater likelihood of conviction if the defendant was charged with more than one crime and if the victim reported the crime within an hour. Conviction was less likely if the victim engaged in risk-taking behavior at the time of the incident and if the offender was black. The offender's age had a negative relationship with conviction.
The data presented in Table 5 indicate that the likelihood of conviction was affected by the number of aggravating circumstances in the case. Defendants in cases with none (or with only one) of the four aggravating factors were less likely to be sentenced to prison than were defendants in cases with three or four aggravating factors. Incarceration was more likely if the defendant was convicted of sexual assault rather than another felony, if the defendant was convicted of more than one offense, if the defendant had a prior felony conviction, and if there was a witness to the incident. Defendants convicted in the post-reform period were less likely to be sentenced to prison than those convicted in the pre-reform period.
These findings suggest that sentence severity in sexual assault cases is determined primarily by legal factors such as the seriousness of the offense and the offender's prior criminal record. The data presented in Table 6 reveal that legal variables also influence the length of the prison sentence. We found that sentence length was affected by the number of aggravating factors (the mean sentence for simple rapes was eighty months less than the sentence for the most aggravated rapes), the seriousness of the conviction charge, the number of conviction charges, and the offender's prior criminal record.
The data displayed in Table 6 also reveal that four extralegal variables influence the length of the sentence. Judges gave significantly shorter sentences to offenders who pied guilty (b=70.80), to offenders convicted of assaulting women who engaged in some type of risk-taking behavior at the time of the incident (b=31.78), and to offenders convicted of assaulting black women (b=46.54). They also imposed significantly longer sentences on black offenders than on white offenders (b=68.74).
Previous research on the impact of rape law reform focused on the reforms' effect on the number of reports of rape and on rates of prosecution, conviction, and incarceration for rape. Although these studies did not find the widespread instrumental results predicted by reformers, none examined the reforms' impact on different types of rape cases.
Our study focused on this issue. Building on Estrich's assertion that traditional rape laws were applied most stringently and most consistently in simple (as opposed to aggravated) rape cases, we suggested that the rape law reforms may have had subtle effects that were masked in earlier studies. More specifically, we speculated that the reforms' impact may have been confined to cases of simple rape.
We found some evidence in support of these suggestions. Our hypothesis that the proportion of simple rape cases bound over for trial would be greater in the post-reform period than in the pre-re-form period was confirmed. A larger proportion of the rape cases reaching the criminal justice system in the post-reform era were cases of rape by unarmed acquaintances who acted alone and did not seriously injure the victim.
There are several possible interpretations of this finding. It is conceivable, although not very' likely, that in the post-reform period there simply are more sexual assaults by unarmed lone offenders who are acquainted with their victims. A more plausible explanation is that the reforms may have prompted more victims of simple rape to report the crime to the police. This seems likely, given our earlier finding that the Michigan reforms produced a significant increase in the number of reports of rape in Detroit. Victims who were reluctant to report the crime to the police in the pre-reform era may have been encouraged to do so in the post-reform period by the widely publicized new laws.
It also is possible that the criteria used by police and prosecutors in screening cases prior to trial changed following the reforms. Changes in the definition of the crime, elimination of resistance and corroboration requirements, and enactment of a strong rape shield law may have made it less likely that police would unfound the complaint, and more likely that prosecutors would file charges, in less serious or "borderline" cases. This would be consistent with the findings from our earlier study. We found that, even with the increase in reporting, the percentage of cases bound over for trial increased in Detroit following the reforms.
Since we do not have data on the characteristics of cases reported to the police during the fifteen year time period, we cannot say conclusively which interpretation is the correct one. Information provided by the criminal justice officials we interviewed in Detroit, however, leads us to speculate that changes in the types of cases bound over for trial reflect both variations in the types of cases reported to the police and changes in police and prosecutor screening criteria. We believe that women raped by unarmed acquaintances were more likely to report the crime to the police following the reforms. We also believe that in the post-reform period police and prosecutors were more likely to regard these simple rapes as real rapes.
Our interpretation of the increase in the proportion of simple rape cases is complicated by the fact that we found no support for our hypothesis that simple rapes bound over for trial would be taken more seriously in the post-reform period. When we compared case outcomes in the two time periods, we found that simple rapes were neither less likely to be dismissed, more likely to result in a conviction, nor more likely to result in incarceration following the reforms. This was confirmed by multivariate analyses incorporating both degree of aggravation and other legal and extralegal predictors of sexual assault case processing decisions. We found no support for our hypothesis that the number of aggravating circumstances would have less impact on case outcomes in the post-reform period than in the pre-reform period.
There are at least two ways to interpret these findings. It may be that criminal justice officials and jurors in Detroit differentiated between aggravated and simple rapes in both the pre-reform and the post-reform period; they may have treated aggravated rapes more seriously in the pre-reform period and continued to treat them more seriously following the reforms. On the other hand, it may be that criminal justice officials and jurors did not differentiate between these types of cases in either time period. Simple rapes, in other words, may have been taken seriously prior to the enactment of rape reform legislation in this particular jurisdiction.
Examination of the results of our multivariate analyses suggests that the first interpretation is more appropriate for the two sentencing decisions, while the second interpretation is more appropriate for the decision to dismiss the charges and the likelihood of conviction.
We found that the number of aggravating circumstances did affect the two sentencing decisions. Offenders convicted of simple rapes were significantly less likely than offenders convicted of rapes with three or four aggravating circumstances to be sentenced to prison; their sentences were also eighty months shorter than the sentences imposed on offenders convicted of aggravated rape.
It is not surprising that aggravated rapes resulted in more severe sentences than simple rapes, and that this did not change following the enactment of rape reform legislation. The seriousness of the offense (i.e., the number of aggravating circumstances) is a legal factor that the judge legitimately takes into account in deciding whether to sentence the offender to prison and in determining the length of the sentence. An offender who uses a gun or knife to sexually assault a stranger, and who seriously injures the victim in the process, will be viewed as more deserving of a long prison sentence than an unarmed offender who sexually assaults, but does not otherwise injure, an acquaintance. The enactment of rape reform legislation would not alter these assessments.
Our findings concerning the dismissal and conviction rates were somewhat different. Degree of aggravation not only did not have differential effects on the likelihood of charge dismissal or conviction in the two time periods, it had no effect on either of these two case outcomes. These results call into question Estrich's assertion that simple and aggravated rapes evoke different responses from criminal justice officials. At least in Detroit, the simple rape cases that survived initial screening decisions by the police and prosector were neither more likely to be dismissed nor less likely to result in a conviction.
One possible explanation for the general lack of support for the hypotheses advanced in this study is that the attitudes of criminal justice officials and the procedures for handling rape cases began to change before the enactment of rape reform legislation. It is possible, in other words, that criminal justice officials in Detroit began to abandon stereotypes concerning "real rapes" and "genuine victims" in the years preceding the 1975 reforms. If this were the case, we would not expect to find significant changes in case outcomes for simple and aggravated rape cases in the post-reform period. Although an analysis of data for a longer pre-reform time period might find evidence of the predicted impact, such a finding would reflect change over time due to changes in attitudes and not change due specifically to the reforms.
Considered together, our findings provide mixed evidence concerning the impact of rape law reform. The fact that the strong and comprehensive Michigan reforms not only did not produce the dramatic instrumental results envisioned by reformers, but did not even produce the more subtle effects investigated here, provides additional evidence that reformers had unrealistic expectations for the rape law reforms. On the other hand, the fact that the proportion of simple rape cases bound over for trial increased significantly in the post-re-form period suggests that more borderline cases are being reported by victims, and accepted by police and prosecutors. It suggests that the rape law reforms have produced a climate more conducive to the full prosecution of cases of simple rape.
* This manuscript is based on work supported by the National Science Foundation under Grant No. SES-9010826. Points of view are those of the authors and do not necessarily represent the position of the National Science Foundation.
1 JEANNE C. MARSH ET AL., RAPE AND THE LIMITS OF LAW REFORM 11-23 (1982); CASSIA SPOHN & JULIE HORNEY, RAPE LAW REFORM: A GRASSROOTS REVOLUTION AND ITS IMPACT 17-25 (1992); Vivian Berger, Man's Trial Woman's Tribulation: Rape Cases in the Courtroom, 77 GOLUM. L. REV. 1, 7-10; Leigh Bienen, Rape III--National Developments in Rape Reform Legislation, 6 WOMEN'S RTS. L. REP. 171, 179-76 (1980).
2 SUSAN ESTRICH, REAL RAPE 47-49 (1987).
3 Id. at 15-26; SPOHN & HORNEY, supra note 1, at 18.
4 Patricia Searles & Ronald J. Berger, The Current Status of Rape Reform Legislation: An Examination of State Statutes, 10 WOMEN'S RTS. L. REP. 25 (1987).
5 Id. at 25-26; ESTRICH, supra note 2, at 81-91.
6 SPOHN & HORNEY, supra note 1, at 23-25.
7 Harriet Galvin, Shielding Rape Victims in the State and Federal Courts: A Proposal for the Second Decade, 70 MINN. L. REV. 763 (1986); Frank Tuerkheiner, A Reassessment and Redefinition of Rape Shield Laws, 50 OHIO ST. LJ. 1245, 1247-50 (1989).
8 MARSH ET AL., supra note 1, at 4-6; Kenneth A. Cobb & Nancy R. Schauer, Legislative Note: Michigan's Criminal Sexual Assault Law, 8 U. MICH. J.L. REF. 184, 217 (1974); Helene Sasko & Deborah Sesek, Rape Reform Legislation: Is it the Solution? 24 CLEVELAND ST. L. REV. 463, 502 (1975).
9 Cobb & Schauer, supra note 8, at 184; Gerald D. Robin, Forcible Rape: Institutionalized Sexism in the Criminal Justice System, in THE CRIMINAL JUSTICE SYSTEM AND WOMEN: OFFENDERS, VICTIMS, WORKERS 241, 256-58 (Barbara R. Price & Natalie J. Sokoloff eds., 1982).
10 Susan Caringella-MacDonald, Sexual Assault Prosecution: An Examination of Model Rape Legislation in Michigan, 4 WOMEN & POL. 65, 71 (1984); MARSH ET AL., supra note 1, at 30-31.
11 Wallace D. Loh, The Impact of Common Law and Reform Rape Statues on Prosecution: An Empirical Study, 55 WASH. L. REV. 544, 613 (1980).
12 Kenneth Polk, Rape Reform and Criminal Justice Processing, 31 CRIME & DELINQ. 191, 195-97 (1985).
13 Karen Gilchrist & Julie Horney, Assessing the Impact of Changes in the Nebraska Rape Statutes: Effect on Prosecution. Paper Presented at the Annual Meeting of the Western Society of Criminology (1980).
14 SPOHN & HORNEY, supra note 1, at 77-105.
15 Id. at 171-72.
16 Cassia Spohn &Julie Horney, Rape Law Reform and the Effect of Victim Characteristics on Case Processing, 9 J. QUANTITATIVE CRIMINOLOGY 383, 397-405 (1993).
17 HARRY KALVEN JR. & HANS ZEISEL, THE AMERICAN JURY 252 (1966).
18 ESTRICH, supra note 2, at 4-7.
19 KALVEN & ZEISEL, supra note 17, at 250; ESTRICH, supra note 2, at 15-26.
20 KALVEN & ZEISEL, supra note 17, at 252-55.
22 ESTRICH, supra note 2, at 10-26.
23 Id. at 29.
24 Id. at 28.
25 Id. at 28-29.
27 MICH. COMP. LAWS ANN. sections 750.520a-750.5201 (Supp. 1980).
28 Caringella-MacDonald, supra note 10, at 66; MARSH ET AL., supra note 1, at 3. For an alternative view, see Ronald J. Berger et al., The Dimensions of Rape Reform Legislation, 22 LAW & SOC'Y REV. 329 (1988).
29 MICH. COMP. LAWS ANN. section 750.520b-e.
30 MICH. COMP. LAWS ANN. section 750.520i.
31 MICH. COMP. LAWS ANN. section 750.520b-e.
32 MICH. COMP. LAWS ANN. section 750.520h.
33 MICH. COMP. LAWS ANN. section 750.520j.
34 MICH. COMP. LAWS ANN. section 750.520j(1).
35 MICH. COMP. LAWS ANN. section 750.520j(1)(a)(b).
36 SPOHN & HORNEY, supra note 1, at 36-37.
37 Id. at 171.
38 MARSH ET AL., supra note 1, at 4-6.
39 Officials in the Sex Crimes Unit explained that they did not have space to store all of the files and that files from the early 1970s were destroyed. They also assured us that there was no pattern to files missing for subsequent years; some files were misfiled, others may have been kept by officers for "sentimental" reasons, and others may have been lost or inadvertently destroyed.
40 Our initial analyses of the data revealed significant differences in the types of cases in the pre- and post-reform periods. Most importantly, we found that there were significantly more cases with child victims in the post-reform than in the pre-reform period. Although some of the defendants in the child victim cases in the post-reform period would have been charged with "rape" under the old statutes, many more would have been charged with indecent liberties with a child, incest, child molesting, or contributing to the delinquency of a minor.
41 Our op-scan form was designed to enable us to create equivalent pre-reform and post-reform samples. The Detroit Police Department records basic information about criminal complaints on an Official Complaint Record (OCR). For sex crimes reported in the post-reform period, the kind of offense is double-coded; the officer investigating the complaint records both the charge under the new statute (i.e., criminal sexual conduct in the first, second, third, or fourth degree) and the charge under the old law (eg., rape, sodomy, incest). Sexual assaults are double-coded because police must still provide the FBI with the number of reports of "forcible rape" for the Uniform Crime Reports. We instructed data collectors to record both offenses.
42 KALVEN & ZEISEL, supra note 17, at 250; ESTRICH, supra note 2, at 4-7.
43 KALVEN & ZEISEL, supra note 17, at 250; ESTRICH, supra note 2, at 4-7.
44 Because there were only 22 cases with all four aggravating circumstances, we combined cases with three or four of the factors.
45 We used the number of charges filed by the prosecutor in analyzing the decisions to dismiss or convict. We used the number of conviction charges in analyzing the two sentencing decisions.
46 THOMAS W. McCAHILL ET AL., THE AFTERMATH OF RAPE (1979); Carol Bobmer, Judicial Attitudes Toward Rape Victims, 57 JUDICATURE 203, 204-06 (1987); KALVEN & ZEISEL, supra note 17, at 254; Gary D. LaFree, Official Reactions to Social Problems: Police Decisions in Sexual Assault Cases, 28 SOC. PROBS. 582, 588 (1981).
47 HUBERT S. FEILD & LEIGH B. BIENEn, JURORS AND RAPE: A STUDY IN PSYCHOLOGY AND LAW 171 (1980); KALVEN & ZEISEL, supra note 17, at 254; McCAHILL, supra note 46, at 105; Shirley Feldman-Summers & Karen Lindner, Perceptions of Victims and Defendants in Criminal Assault Cases, 3 CRIM. JUST.& BEHAV. 135, 145 (1976); Lisa Frohmann, Discrediting Victims' Allegations of Sexual Assault: Prosecutors' Accounts of Case Rejections, 38 SOC. PROBS. 213, 218 (1991); Barbara F. Reskin, & Christy A. Visher, The Impact of Evidence and Extralegal Factors in Jurors' Decisions, 20 LAW & SOC'Y REV. 427, 431 (1986).
48 McCAHILL ET AL., supra note 46, at 191.
49 GARY D. LAFREE, RAPE AND CRIMINAL JUSTICE: THE SociAL CONSTRUCTION OF SEXUAL ASSAULT 99-111 (1989); Cassia Spohn et al., The Effect of Race on Sentencing: Re-Examination of an Unsettled Question, 16 LAW & SOC'Y REV. 71, 80-84 (1981-1982).
50 As noted above, Aggrp3 is the omitted category in all of the analyses.
51 ESTRICH, supra note 2, at 29.
52 SPOHN & HORNEY, supra note 1, at 86.
54 See supra Table 6.
55 ESTRICH, supra note 2, at 29.
56 Id. at 6.
GRAPH: Figure 1. A COMPARISON OF SEXUAL ASSAULT CASES IN PRE- AND POST-REFORM PERIODS
Variable Code N % Dependent Variables All charges dismissed 0=no 653 80.4% 1=yes 159 19.6 Found guilty at trial[a] 0=no 129 42.7 1=yes 173 57.3% Convicted (plea or trial)[b] 0=no 287 35.4 1=yes 523 64.6 Sentenced to prison[c] 0=no 146 27.8 1=yes 379 72.2 Mean sentence (in months) Mean = 182.22 Independent Variables Case Characteristics Scale Four dummy variables; Aggrp3 of aggravation[d] is the omitted category Aggrp0 -- 179 22.0 Aggrp1 -- 225 27.7 Aggrp2 -- 268 33.0 Aggrp3 -- 140 17.2 Number of charges Mean = 2.53 Most serious conviction Three dummy variables; other charge[e] felony is the omitted category Sexual assault -- 308 58.7 Other sex offense -- 103 19.6 Other felony -- 109 20.8 No. of conviction charges Mean = 1.50 Crime involved vaginal-penile penetration 0=no 129 15.9 1=yes 683 84.1 Offender used threats or force 0=no 132 16.3 1=yes 680 83.7 Witness to incident 0=no 631 78.8 1=yes 170 21.2 Physical evidence of crime[f] 0=no 272 33.5 1=yes 540 66.5 Defendant pled guilty 0=no 302 46.2 1=yes 351 53.8 Pre- or Post-reform case[g] 0=pre 279 34.4 1=post 533 65.6 Victim Characteristics Race 0=white 210 26.2 1=black 593 73.8 Age Mean = 23.0 Evidence of risk-taking behavior[h] 0=no 505 62.2 1=yes 307 37.8 Screamed during attack 0=no 628 77.3 1=yes 184 22.7 Physically resisted offender 0=no 485 59.7 1=yes 327 40.3 Reported crime within 1 hour 0=no 392 49.7 1=yes 396 50.3 Offender Characteristics Race 0=white 91 11.3 1=black 714 88.7 Age Mean = 26.2 Prior felony conviction 0=no 558 68.7 1=yes 254 31.3
a Includes only defendants who were tried by a judge or jury.
b Includes all defendants. Variable coded "1" if defendant pled guilty or was found guilty at trial and "0" if the charges were dismissed or the defendant was acquitted.
c Percentage of convicted defendants who were sentenced to prison.
d Scale of aggravation = stranger (victim and offender strangers) + weapon (offender used a gun or knife) + injury (offender injured the victim) + more than one offender. Scale ranges from 0 (none of the four aggravating circumstances) to 3 (3 or 4 of the aggravating circumstances).
e All of the defendants in the sample were charged with sexual assault (rape, sodomy and gross indecency in the prepreform period; criminal sexual conduct in the first degree or criminal sexual conduct in the third degree in the post-reform period), but defendants were convicted of a variety of offenses. Other sex offenses = criminal sexual conduct in the second degree, attempted raped, attempted criminal sexual conduct, or assault with intent to rape. Other felony = assault, robbery, and other felony offenses.
f Includes semen, fingerprints, blood stains, hair or skin samples.
g The data file included cases bound over for trial from 1970 through 1984. Cases from January of 1970 through March of 1975 were classified as pre-reform cases. Cases from April of 1975 through December of 1984 were classified as post-reform.
h Variable coded "1" if police file indicated that at the time of the incident the victim was walking alone late at night, was hitchhiking, agreed to accompany offender to residence, invited offender to own residence, was in a bar alone, was using alcohol, or was using drugs.
Legend for Chart: A - Degree of Aggravation B - Charges Dismissed: Pre C - Charges Dismissed: Post D - Found Guilty At Trial: Pre E - Found Guilty At Trial: Post F - Convicted-By Plea or Trial: Pre G - Convicted-By Plea or Trial: Post H - Sentenced To Prison: Pre I - Sentenced To Prison: Post J - Mean Sentence (In Months): Pre K - Mean Sentence (In Months): Post A B C D E F G H I J K 0 .265 .246 .167 .500[*] .531 .554 .462 .500 75.07 112.89 (49) (130) (12) (52) (49) (130) (26) (72) (14) (38) 1 .156 .169 .609 .523 .727 .626 .643 .656 121.40 160.84 (77) (148) (23) (65) (77) (148) (56) (93) (43) (69) 2 .187 .224 .567 .644 .692 .644 .878 .769 166.81 277.90[*] (107) (161) (30) (59) (107) (161) (74) (104) (59) (79) 3 .152 .149 .529 .750 .674 .734 .839 .913 174.12 234.33 (46) (94) (17) (44) (46) (94) (31) (69) (26) (55)
Variable b SE Scale of Aggravation Aggrp0 0.20 .36 Aggrp1 -0.20 .34 Aggrp2 0.13 .30 Case Characteristics No. of charges -0.18 .07[*] Crime involved vaginal-penile penetration -0.14 .28 Offender used threats or force 0.40 .29 Witness to incident 0.25 .23 Physical evidence of crime 0.27 .22 Post-reform case 0.23 .23 Victim Characteristics Race=black -0.52 .26[*] Age -0.02 .01 Evidence of risk-taking behavior 0.34 .20 Screamed during attack -0.57 .25[*] Physically resisted offender -0.13 .20 Reported crime within one hour -0.45 .20[*] Offender Characteristics Race=black 0.98 .40[*] Age 0.02 .01 Prior felony conviction -0.01 .22 No. of cases 769 -2 log-likelihood X[sup 2] = 711.29 df = 750 P = 0.8413 Goodness of fit X[sup 2] = 775.65 df = 750 P = 0.2508 Effect of adding aggrp x reform interaction terms Improvement X[sup 2] = 0.872 df = 3 P = 0.8322
* P </= 5.05
Variable b SE Scale of Aggravation Aggrp0 -0.29 .30 Aggrp1 0.11 .27 Aggrp2 0.04 .26 Case Characteristics No. of charges 0.27 .06[*] Crime involved vaginal-penile penetration -0.29 .24 Offender used threats or force -0.11 .23 Witness to incident 0.15 .20 Physical evidence of crime 0.20 .18 Post-reform case -0.36 .19 Victim Characteristics Race=black 0.10 .23 Age 0.01 .008 Evidence of risk-taking behavior -0.46 .17[*] Screamed during attack 0.22 .20 Physically resisted offender 0.02 .17 Reported crime within one hour 0.45 .17[*] Offender Characteristics Race=black -0.75 .33[*] Age -0.02 .01[*] Prior felony conviction -0.11 .18 No. of cases 766 -2 log-likelihood X[sup 2] = 930.87 df = 748 P = 0.0000 Goodness of fit X[sup 2] = 765.66 df = 748 P = 0.3191 Effect of adding aggrp x reform interaction terms Improvement X[sup 2] = 0.856 df = 3 P = 0.8361
* P </= .05
Variable b SE Scale of Aggravation Aggrp0 -1.04 .49[*] Aggrp1 -0.86 .42[*] Aggrp2 -0.25 .46 Case Characteristics Most serious conviction charge Sexual assault 0.70 .32[*] Other sex offense -0.10 .35 No. of conviction charges 1.54 .40[*] Crime involved vaginal-penile penetration -0.62 .35 Offender used threats or force 0.16 .33 Witness to incident 0.68 .33[*] Physical evidence of crime -0.05 .27 Offender pied guilty -0.49 .30 Post-reform case -0.94 .29[*] Victim Characteristics Race=black -0.54 .38 Age 0.026 .014 Evidence of risk-taking behavior -0.19 .27 Screamed during attack -0.51 .29 Physically resisted offender -0.22 .27 Reported crime within one hour -0.02 .25 Offender Characteristics Race=black 0.25 .47 Age -0.03 .02 Prior felony conviction 1.24 .34[*] No. of cases 485 -2 log-likelihood X[sup 2] = 424.56 df = 463 P = 0.8994 Goodness of fit X[sup 2] = 411.09 df = 463 P = 0.9601 Effect of adding aggrp x reform interaction terms Improvement X[sup 2]= 6.919 df = 3 P = 0.0745
* P </= .05
Variable b Beta t Scale of Aggravation Aggrp0 -80.20 -.16 2.86[*] Aggrp1 -31.94 -.09 1.40 Aggrp2 33.01 .10 1.59 Case Characteristics Most serious conviction charge Sexual assault 58.23 .17 2.96[*] Other sex offense -26.36 -.06 1.08 No. of conviction charges 39.09 .25 4.91[*] Crime involved vaginal-penile penetration -4.83 -.01 0.22 Offender used threats or force -33.48 -.07 1.44 Witness to incident -12.85 -.03 0.71 Physical evidence of crime -12.19 -.03 0.71 Offender pled guilty -70.80 -.20 4.27[*] Post-reform case -2.46 -.01 0.14 Victim Characteristics Race=black -46.54 -.13 2.44[*] Age -0.61 -.04 0.96 Evidence of risk-taking behavior -31.78 -.09 1.97[*] Screamed during attack -16.13 -.04 0.90 Physically resisted offender 28.72 .08 1.75 Reported crime within one hour 24.49 .07 1.59 Offender Characteristics Race=black 68.74 .13 2.53[*] Age 1.18 .05 1.07 Prior felony conviction 38.28 .11 2.24[*] No. of cases = 358 R[sup 2] = .39 Effect of adding aggrp x reform interaction terms R[sup 2] = .40 R[sup 2] Change = .005 F Change = 0.9347 Sig. = .4241
a Sentence (in months) imposed on those who were sentenced to prison.
* P </= .05
By CASSIA C. SPOHN and JULIE HORNEY
Cassia C. Spohn is a professor of criminal justice at the University of Nebraska at Omaha. She is the co-author of two books: The Color of Justice: Race, Ethnicity, and Crime in America (with Sam Walker and Miriam DeLone); and Rape Law Reform: A Grassroots Movement and Its Impact (with Julie Homey). Her research interests include the effect of race and gender on court processing decisions, victim characteristics and case outcomes in sexual assault cases, and the impact of rape law reform. She is currently the principal investigator for multi-jurisdiction study of the effect of race on sentencing. Julie Homey is Professor of Criminal Justice at the University of Nebraska at Omaha. In addition to case processing decisions, her current research interests include individual patterns of offending, the situational correlates of violence, and the sensitivity of individual offenders to social controls.