Skip to main content
Photo of a Law Library

Normative Words and the Fundamental Limits of Rape Law Reform

Eric R. Carpenter[1]*

Most rapists will not be held accountable. For every 100 forcible rapes, less than six offenders will be convicted. Since the 1970s, legislatures have been reforming sexual assault laws to hold more offenders accountable, but as this Article demonstrates through a comprehensive review of the research, the reforms do not appear to have worked. This Article then brings together social science, law, and practice to argue that this is because normative words in the law allow rape myths to enter the legal system. These words existed in the pre-reform laws and still exist in the post-reform laws. They remain within the consent element (which is governed by the Confrontation Clause) and the mistake of fact defense (which is governed by the Due Process Clause). Both the Confrontation Clause and the Due Process Clause are fundamentally normative and fundamentally fixed, so normative words will always be in rape law and will always serve as a potential entry point for bias. Faced with these fundamental limits on rape law reform, this Article further provides recommendations for reforms that may help keep some bias from entering the legal system and concludes that the way to improve case processing is to ensure that law enforcement and prosecutors operate free of inaccurate generalizations about rape.


Most rapists will not be held accountable. For every 100 forcible rapes, only 5–20% will be reported, only 0.4–5.4% will be prosecuted, only 0.2–5.2% will result in a conviction, and only 0.2–2.8% will result in incarceration.[2] The feminist critical theory that explains this high rate of case attrition is straightforward: legal actors may have inaccurate beliefs about what rape looks like (these beliefs are called rape myths); consent-defense rapes (meaning non-stranger, non-familial) don’t fit that image; and so legal actors may undervalue the cases. Law enforcement officers and prosecutors may drop the cases, and judges and jurors may acquit the offender if the case makes it to trial.[3]

Reformers recognized that the common law incorporated rape myths, making it easy for these beliefs to enter the system.[4] As a result, many jurisdictions abandoned the common law for reform models; but when researchers studied whether the reforms affected case processing, they could not find persuasive evidence that the reforms worked as intended.[5] The reason that these legal changes have not impacted case processing could be due to the impact of normative words. These words—words like reasonable, should, and sufficient—are the entry points for rape myths. These existed in the common law before the reforms, and they still exist in the law after the reforms.[6]

Moreover, these entry points for rape myths will always be there. This is because normative words exist in two areas of the law that either will not be reformed or cannot be reformed: the evaluation of victim credibility associated with the consent element and the mistake of fact defense.[7] First, victim credibility is governed by the Confrontation Clause—which is fundamentally normative and fundamentally fixed.[8] Next, it is unlikely that jurisdictions will eliminate the mistake of fact defense, and that defense is governed by the Due Process Clause—which is fundamentally normative and fundamentally fixed.[9] While some reforms may reduce the entry points for rape myths, they can only do so much. The true fix for changing case processing is changing norms before they arrive at the entry points in the law. Lamentably, changing norms at the macro or societal level takes a long time. But with the proper training of law enforcement and prosecutors, changing norms at the micro or trial level can happen now.

This Article makes a streamlined and simplified argument by bringing together sources from social science, law, and practice. Many sections could support (and have supported) separate, full-length articles. The primary purpose of this Article is to keep the argument in sight by generalizing legal models when possible, speeding up in sections that have robust treatment in the literature, and slowing down in sections that do not. Part II describes the reasoning patterns associated with rape myths, how the common law endorsed these myths, and how the reform models were designed to address them. Part III provides a thorough survey of the research where social scientists tried to measure whether the reforms had any effect on case processing. The results are inconclusive and inconsistent but suggest that reforming the law does not have any major effect beyond improvements in victim reporting. Part IV argues that the failure of rape law reform is because of the normative words that exist in the reform models, which will not or cannot be removed. Faced with these fundamental limits on rape law reform, Part V offers suggestions for reforms that may be of some use. Part VI concludes by proposing that the way to improve case processing is to ensure that law enforcement and prosecutors operate free of inaccurate generalizations about rape and to have prosecutors educate judges and jurors, case by case, on the realities of rape.

The Reform Effort

Reasoning Patterns Associated with Rape Myths

According to feminist critical theory, gender bias causes case attrition in consent-defense cases.[10] This bias comes from gender role beliefs, specifically, the acceptance of rape myths.[11] Rape myths are attitudes and beliefs about rape that “are generally false but are widely and persistently held, and that serve to deny and justify male aggression against women.”[12] These include beliefs that only deviant men rape; men cannot control their sexual urges; the woman wanted it or deserved it; women lie about rape; no harm was done; or that certain events do not qualify as “real” rape.[13] These generalizations or schemas are inaccurate and cause legal actors to then improperly devalue the cases.[14]

The central rape myth is an event schema (or generalization) about what “real” rape looks like. “Real” rape involves a deviant man who uses violence and weapons against a woman who is a stranger, causing injuries in the process.[15] Within that event schema, there are particular person schemas about rapists and victims. “Real” rapists come from outside of your social group, known as the outgroup.[16] These men look different, are violent, and are sexual deviants.[17] “Real” victims come from within your social group (the ingroup) and are blameless in the assault.[18] They are attacked by surprise in a parking lot or while jogging.

Looking at the generalization that ingroup men don’t rape, that belief has some truth to it. The overwhelming majority of men (ingroup and outgroup) do not rape—only about 5–10% of men commit sexual assaults.[19] And it turns out that these men are deviant, but not in the ways projected by rape myths. This 5–10% who commit rapes tend to be narcissistic, aggressive, hostile to women, hyper-masculine, and lacking in empathy.[20] These men identify victims who are vulnerable and who may not be believed if they report; they try to isolate their victims; they use only the force necessary to commit the offense (which could often be just pinning down a part of the victim’s body); they use psychological weapons and surprise; and they use alcohol as a weapon.[21]

We can understand perpetrators’ deviancy better when we contrast it with what we know about victim behaviors. Mary Koss and colleagues have found that in non-stranger assaults, 83% of the victims reasoned or pled with the man (probably saying things like “no” and “I don’t want to do this” and “stop”), 76.6% of the women turned cold, 69.6% struggled, 45.7% cried or sobbed, and 11.2% screamed or tried to run away.[22] These men force sex on women who are pleading, turning cold, and crying. Men with normal sexual arousal patterns do not find that type of behavior from a sexual partner to be stimulating; however, the men in this 5–10% percent do.[23]

The case processing problem is that this 5–10% percent are good at disguising their deviancy.[24] When one of these men is sitting in front of police officers or jurors, the legal actors use their generalizations about ingroup men and fail to see the person that the victim saw during the assault, who in a moment went from being a person that the victim thought she knew to a stranger that forced sex on her.[25] The legal actors do not see the person that ignored the victim’s pleas to stop and her crying and who pinned her down with his body weight.

If a legal actor believes that only deviant men rape, then when a woman reports an assault against an ingroup man, the legal actor experiences dissonance. Ingroup men don’t rape, so something else must be going on. If the woman appears to be outgroup (based on how she behaves or dresses), the calculus is pretty easy. According to rape myths, outgroup women often lie about rape to protect their reputations or their relationships (if it looks like they were cheating) or because they are vindictive.[26] This ingroup man didn’t rape her because she actually consented.[27] Plus, outgroup women often deserve what is coming to them based on how they behave and dress.[28]

If the woman does appear ingroup, the legal actor still needs to resolve why an ingroup woman would accuse this ingroup man of rape. To resolve the dissonance, the legal actor may think that the ingroup woman did consent, but the legal actor may decide that, unlike what we might expect from an outgroup woman, this ingroup woman isn’t a vindictive liar.[29] Instead, the ingroup woman is simply confused because her friends told her that her experience looked like rape, or she could be unconsciously rewriting the events in her mind to preserve her self-esteem (“because she regrets the sex or feels cheap”) and to preserve her place in the ingroup.[30] The legal actor may conclude this was drunken sex followed by regret, with a consensual sex act recast by the victim as rape.[31]

If the legal actors do believe the woman, then they still have dissonance: she was raped, but not by a rapist.[32] Ingroup men don’t rape, so the legal actors need another reasoning pathway to solve the problem: this ingroup man must have been mistaken that this ingroup woman consented.[33] This was just a miscommunication. He must have thought she had consented, probably because she did not adequately communicate that she did not want to have sex. This belief that ingroup men misunderstand women because women send unclear signals is called the Miscommunication Hypothesis.[34] Research shows that this hypothesis is false,[35] making it what Thomas MacAulay Miller calls the “mythcommunication.”[36] Instead, women communicate consent and lack of consent in ways that often are subtle—but which are at the same time unambiguous.[37] Women scoot away, move the man’s hand, do not reciprocate the acts, suggest that the couple do something else, give disapproving looks, and say that they have to get up early, among other things.[38]

Men fully understand these communications.[39] Just as in regular social situations where people say “no” in subtle ways (“Oh, I really would like to, but I already made plans”), and just as people in regular social situations understand that the message is “no,” so do people in sexual situations. In sexual situations, the vast majority of men hear the rejection and stop. They say, “Huh? What’s wrong? I thought things were going well,” and this normal behavior makes it clear that there is no consent.[40] About 5–10% of men also hear the “no” but then decide that they are entitled to sex and force sex acts on the victim.[41] For this 5–10%, “[i]t’s not that they don’t understand, they just don’t like the answer.”[42]

These various reasoning patterns enter the decision-making process through legitimate decision-making factors.[43] Generally, a case is more likely to make it through the system if the victim is physically injured or a weapon is used, or if the evidence against the suspect is strong (where the strength of the evidence is often measured by the victim’s willingness to participate, the availability of other witnesses, and the availability of forensic evidence).[44] While those factors appear legitimate, unsurprising, and free of explicit bias, research suggests that these factors serve as the entry points for latent bias.[45] When they are present in a case, they tend to make the victim appear more credible and cast the case within the generalized image of what a rape case is supposed to look like—an image that is itself shaped by rape myths.[46] When those factors are present, the case is also more attractive for law enforcement and particularly for prosecutors because they may be more likely to get a conviction.[47]

Conversely, when these factors are not present, legal actors may think that the victim is not credible because her story does not fit the legal actors’ narrative of what a rape case should look like, so the legal actors will devalue the case.[48] When the legal actors also see that the victim behaved in ways before, during, or after the assault that are inconsistent with their expectations about how women should behave before, during, or after the assault, they may presume the victim is lying, mistaken, or confused.[49] And if there are discrepancies in the victim’s statements, they may conclude she is lying or not credible.[50]

The influence of rape myths may enter at the earliest stages of legal processing and taint everything else that follows.[51] As Deborah Tuerkheimer has noted, when law enforcement first receives a report of a sexual assault, they may reverse their normal investigative presumptions: “The typical law enforcement investigation is guilt-presumptive . . . . In sexual assault cases, this presumption is flipped. Investigators start from the proposition that the complainant is lying and act to confirm this belief.”[52]

We see this shift in presumptions in official statements made by the Army’s law enforcement leadership in Generating Health & Discipline in the Force.[53] There, leadership reports (incorrectly) that 41–50% of sex assault reports are false allegations.[54] This is a warning to investigators not to believe victims. Nowhere else in this report—a lengthy report that covers a wide range of crimes—does the leadership suggest that victims are untrustworthy.[55] The leadership then builds on this warning by telling investigators that because the false reporting rate is so high, investigators need to take special care of the subjects of these investigations.[56] They tell investigators that they should not infer that an accused is guilty when they receive an allegation of sexual assault (reversing the normal presumption) and to protect and balance the rights of the victim and the alleged offender.[57] Nowhere else in this report—which is focused on prosecuting offenders—do they tell investigators to be non-adversarial with the offenders of other crimes.[58]

That flipping of the investigative presumption may be the initial moment of bias.[59] Because law enforcement has flipped the ordinary presumptions, law enforcement officers may compile weak files.[60] Investigators may then fail to check the offender’s background to see if he has been accused of something similar before, and they may not look for evidence and witnesses that corroborates the victim’s account.[61] These weak case files may then influence the other legal actors that are downstream.[62] Further, law enforcement treatment of the victim, especially with hostile interrogations, may cause the victim to disengage from the case and ultimately end it.[63] Victims, knowing the treatment they are about to endure and facing very real prospects that they will endure that treatment only to have the case dropped anyway, may opt out of the system and not report.[64]

Rape Myths and the Common Law

These reasoning patterns fit neatly into the legal framework that existed before the reform movement began in earnest in the 1970s.[65] The existing Anglo-American common law incorporated rape myths, and these myths then entered procedural rules, evidentiary rules, and the substantive law as the common law transitioned into code.[66] For example, procedural rules required prompt complaints.[67] If the victim delayed making a report for whatever reason, she formally lost access to the legal system.[68] Such procedural rules endorsed the belief that women routinely lie about being raped (if she had been raped as she said, she would have reported right away—she is only making this report now because she is being vindictive). Evidentiary rules required the victim’s testimony to be corroborated (legitimizing the belief that women routinely lie) and allowed opinion and reputation testimony about the victim’s character for chastity (legitimizing the belief that a woman like her asked for it or deserved it).[69] The substantive law contained spousal exceptions so that husbands could not be convicted of raping their wives,[70] legitimizing the belief that marital rape is not “real” rape.

Further, the basic legal elements of the substantive law incorporated rape myths.[71] The basic elements of common law rape are penile-vaginal intercourse by a man of a woman, without her consent, and by force.[72] This is a general intent crime, where the offender’s culpable mental state is not written into the statute;[73] however, the victim’s mental state is—consent. First, because consent was written into the statute, the prosecutor had the burden of producing evidence during its case-in-chief on that element.[74] Prosecutors do that with direct testimony from the witness about her mental state and with circumstantial evidence that corroborates that testimony.[75] Because the written elements of the statute include the victim’s mental state but not the offender’s mental state, part of the prosecution’s case had to focus on the victim’s behavior and her mental state, rather than the offender’s.[76] This then directs attention during cross-examination to topics that are influenced by rape myths. She must have wanted it (and so really did consent) if she dressed like that, went to a bar, got drunk, danced with several men, and was known to go home with strangers. She must be lying on the stand right now because of buyer’s remorse or because she is a woman scorned.

Writing the element of consent into the statute is not necessary.[77] Many crimes (theft or assault and battery, for example) often do not include an affirmative, written element of consent.[78] In those cases, the prosecution does not have the initial burden of producing evidence of consent, and so the law does not force the prosecution to focus on what the victim did; rather, the prosecution focuses on what the offender did. In those cases, the law requires the defense to raise consent as an issue.[79] The defense has the burden of production, and once met, then the government must prove a lack of consent beyond a reasonable doubt.[80]

Second, lack of consent was not enough for the non-consensual act to be rape. A woman could plainly say “no,” but if the man did not apply a certain amount of force in addition to that refusal, no rape occurred. Critics have pointed out that in other areas of law, a simple “no” is enough.[81] Imagine that John wants to take Frank’s property, and Frank says “No.” If John takes it anyway, he is guilty of theft. John does not also have to apply force when taking it (turning theft into robbery). The common law respects a man’s autonomy over his property but does not respect a woman’s autonomy over her body.[82] This legitimizes the belief that if there is no force, then it was not “real rape,” and so no harm was done, or she must have wanted it—otherwise, she would have stopped him.

Force was then defined by how much the woman resisted (in some jurisdictions, to the utmost; in others, such reasonable measures of resistance as required by the circumstances).[83] Unless a woman fought back with some amount of violence, she would not overcome the strong legal presumption that even though she said “no,” she meant “yes.”[84] Under this legal scheme, the starting point is that woman has consented unless she fights back, rather than a woman has not consented until she says so. The law said “yes, unless” rather than “no, until.” This presumption of consent legitimized the belief that women want it—otherwise, they would have fought back. Notice again that under this scheme, the prosecutor must focus on the reasonableness of the victim’s resistance behavior, not the offender’s entitlement behavior.

Even if the prosecution proved that the victim did not consent, the defense could still raise the defense of mistake of fact as to that consent.[85] Under this defense, the defendant has the burden of producing evidence that he honestly believed that the victim consented, and if he did, that that belief was reasonable under the circumstances.[86] The mistake of fact defense, as is the case with other general intent offenses, acts as a measure of the offender’s culpable mental state.[87] The offender is culpable if he should have known (in other words, was negligent) that an attendant circumstance existed (here, the victim’s consent or lack thereof).[88] Because a mental state is not written into the statute as an element, the other elements of the offense let the fact-finder presume that the offender had a culpable mental state if those elements are proven (here, the accused committed a sex act by force and without consent).[89] The accused can then raise a mistake of fact defense to say that, even though those other elements were proven, he did not have a blameworthy state of mind because he did not and should not have known that one of the attendant circumstances listed in the offense existed. Once raised, the government then has the burden of disproving the defense beyond a reasonable doubt.[90] The government must now prove that the victim manifested indications of consent such that the offender was negligent in ignoring them.[91]

The mistake of fact defense is not unique to rape law, but here it legitimized societal beliefs that ingroup men do not rape (boys will be boys, how was he to know she meant “no?”). And remember, the common law—through the definition of force—has already said when it was reasonable for a man to be mistaken about consent.[92] If the woman does not fight back, then she has not appropriately communicated that she does not want to have sex, and so it is reasonable for a man in that situation to mistake her “no” as “yes.” The focus is still on the victim’s resistance behavior, not the offender’s culpable behavior.

Reform Efforts

Beginning in the 1970s, jurisdictions began to reform these rules and laws, and while “the reforms and laws enacted vary between the states, they all seek to shift the focus away from the victim and toward the behavior of the defendant.”[93] Jurisdictions universally reformed evidentiary and procedural rules.[94] The formal requirements for corroboration and fresh complaints, as well as the spousal exemption, have disappeared.[95] Jurisdictions also adopted various rape shield rules that are supposed to prevent evidence about the victim’s sexual character or previous sexual acts from being introduced at trial unless that evidence is constitutionally required for the defense or meets certain other exceptions.[96]

Looking at the substantive law, jurisdictions have converged in their response to some elements of the crime, like expanding the definitions of a penetrative act beyond penile-vaginal penetration and making the statutes gender neutral.[97] But they have diverged in the primary definition of the crime, with some keeping the basic common law form and others adopting reform models.[98] As a reminder, the elements of the common law model are vaginal penetration by a penis, without consent, and by force.[99]

The main difference in the various reforms to the substantive law is the choice to drop either the force element or the consent element.[100] The reforms tend to fit into three models: the force-centric model (typified by Michigan[101]); a variation of that model, the assault-plus model (typified by Canada[102]); and the consent-centric model (typified by Florida).[103] In the force-centric model, the statute drops the consent element and focuses on the force used by the offender, where force is often no longer formally defined in terms of the woman’s resistance, although shadows of the consent element remain.[104] The Michigan statute is force-centric,[105] and its implementation has been studied.[106] There, the worst offense (first-degree criminal sexual conduct, a felony) is a penetrative sex act with a high degree of force (or other aggravating circumstance);[107] the second-worst (second-degree, a felony) is a nonpenetrative sex act but high-force (or other aggravating circumstance);[108] the third-worst (third-degree, a felony) is penetration but lower-force (or less severe aggravating circumstances);[109] and the fourth-worst (fourth-degree, a misdemeanor) is nonpenetrative and lower-force (or less-severe aggravating circumstances).[110]

In this force-centric model, the prosecution no longer has the burden of producing evidence in its case-in-chief of the victim’s lack of consent.[111] This shifts the initial focus of the prosecution’s case to the offender’s actions rather than the victim’s behavior and mental state.[112] However, the defense can still raise the issue of consent, and in non-stranger, non-familial sexual assaults, consent will almost always be raised.[113] The reform only affects who has to raise the issue of consent and when it will be raised. Further, the defense can still raise the mistake of fact defense.[114]

A variation of the force-centric model is the assault-plus model.[115] This model starts with assault as the baseline offense and then adds sexual contact as an aggravator to the underlying assault. As with the Michigan model, the initial focus is on the offender’s forceful, offensive contact—the assault. In many jurisdictions, the crime of assault (or battery) does not include an express, written element of non-consent.[116] Rather, it only requires a voluntary act by the offender that causes offensive bodily contact or harm to the victim.[117] As with the force-centric model, the assault-plus model allows a defendant to raise the issue of consent or mistake of fact as to consent.[118] This model only differs from the force-centric model in its naming convention. By calling the offense an assault, the legislature hopes to frame the problem as a violent assault and leave behind some of the rape myths that are associated with the word “rape.”[119]

Turning to the consent-centric model, this model drops the force element from the primary offense and focuses on consent.[120] These are “no, unless” statutes, colloquially known as “affirmative consent” statutes.[121] In these, a sex act without consent is a crime.[122] The Florida statute is a good example. The baseline offense, sexual battery, is a second-degree felony.[123] A penetrative act without consent is sufficient for a conviction.[124] From there, degrees of force or other aggravating conditions increase the crime to a first-degree felony or a life felony.[125] Because consent is an element of the offense, the government has the burden of proving a lack of consent in its case-in-chief.[126] In addition, the defense can raise a mistake of fact defense as to consent.[127] In this model, the legislature has decided that the normative expression about a woman’s autonomy—“no, unless”—outweighs the benefit of forcing the defense to raise the issues related to consent.[128]

Some jurisdictions use a hybrid model, giving the government force-centric options that do not include a consent element,[129] while including at least one offense that is consent-centric.[130] Congress chose a hybrid model for the military that went into effect in 2007, and this change has been studied.[131] The Canadian statute is another hybrid, with a mix of an assault-plus statute and a consent-centric model.[132] In Canada, lack of consent is a written element in all assaults,[133] and for sexual assault—like the consent-centric model—affirmative consent is required before the sex act.[134] The change to the definition of sexual assault in Canada was implemented in 1983 and has also been studied.[135] Another important feature of the current Canadian statute is that it narrows the mistake of fact defense by telling us that certain situations amount to an unreasonable mistake of fact.[136] The accused’s mistaken belief is unreasonable if it is due to his intoxication, reckless conduct, or willful blindness.[137] And the defense is unavailable if he did not take reasonable steps to determine if the victim was consenting or when there is no evidence that the complainant’s voluntary agreement to the activity was affirmatively expressed by words or behavior.[138]

All that said, it is important to note that the differences between the models are modest. The difference between the common law and the force-centric model is just timing. In the common law model, the government must present evidence of consent in its case-in-chief. In the force-centric model, the defense raises consent in its case. In both instances, the burden of proof on consent is still on the government. The difference between the common law and the consent-centric model is that the government can get a conviction for the baseline offense without first having to prove force. If the government tries to prove force to get a conviction of a more serious offense, then this model is similar to the common law for that more serious offense. And the mistake of fact defense is available for all of the models.

Measuring the Effects of Rape Law Reforms

Once legislatures passed these reforms, researchers tried to measure whether these changes had any effect on rape case attrition. To date, the research is inconsistent and generally inconclusive. Researchers focus on six decision moments in the life cycle of a case: the victim’s decision to report; law enforcement’s decision to found the case; law enforcement’s decision to clear the case; the prosecutor’s decision to take the case to trial; the judge or jury’s decision to convict; and the sentencing authority’s decision on sentencing.[139]

The basic research problem has two parts. The first is to see if there is a change in processing following the legal reform.[140] The second is to control for other factors that may have caused that change other than the legal intervention.[141] One of those other factors could be some other abrupt mechanism other than the rape law reform (like an abrupt reform that affects all crime). Researchers call this the “history threat,” where “[o]ther events occurring at about the same time actually may be responsible for the effects noted.”[142] One way to account for this factor is to use other similar crimes as a control group to see if there was also a change in the case processing of those crimes.

Another factor could be longer-term changes in norms related to sexual assault or crime in general. The legal reform may not be the causal factor: “In the case of rape law reforms, increased national attention to the problems surrounding the prosecution of rape cases might have sensitized criminal justice officials and led to observed changes in processing.”[143] Instead of having a direct impact on case processing, the changing law may have a distal role, serving more of a ceremonial function that may lead others to change their norms, but the changing norms are the direct causal factor for longer-term change in processing over time.[144]

To control for longer-term trends, researchers can use time-series analysis, which is a statistical process where the model “pre-whitens” the data by accounting for variation that is dependent on prior observations.[145] This dependent variation is likely due to long-term ongoing processes that also occurred during the observed period.[146] In sexual assault cases, that would be longer-term changes in societal norms. Once the data is pre-whitened, the researchers then look for before and after differences.[147]

Detecting a Processing Change

Below are the studies that have looked for a processing change without controlling for other factors. These studies indicate that victim reporting has increased after legal changes, but the findings related to the processing of rape cases beyond that are mixed. Studies have found an increase in cases bound over for trial, the ratio of indictments to reported cases, and conviction rates, but some studies have not found changes in dismissals, the ratio of convictions to indictments, or punishments. Further, we are limited in what we can learn from these studies because we do not know if these trends are unique rape law reforms or if they reflect trends found throughout criminal law processing or other long-term trends.

To start, Spohn and Horney looked at the reforms in 1974 to the Michigan system, using randomly selected files from Detroit for 1970–1984.[148] They found that more simple rape cases were bound over for trial after the reform than before the reform, but they could not find a change in dismissal rates.[149] Next, Gunn and Linden looked at the 1983 reforms to the Canadian system using case files from Winnipeg, Manitoba, for 1981–1982 and 1984–1985 and found that victim reporting went up by 66% but that the reforms had a minimal impact on conviction rates.[150] LeBeau studied California’s changes in 1974 to evidentiary rules. He used data from San Diego for 1971–1975 and reported descriptive statistics that showed an increase in reporting by victims who were casual acquaintances of their assailants.[151] Ajzenstadt and Steinberg studied the 1988 changes to Israel’s punishment scheme. Looking at court records from 1985–1991, they found no major differences in the punishments imposed after the reforms.[152] Next, Loh studied the changes made in 1975 to Washington’s substantive definitions and penalty structure. Looking at case files from 1972–1977, he reported descriptive statistics that conviction rates for rape increased but charging rates remained the same.[153]

Finally, Clay-Warner and Burt used data from national surveys conducted in the United States and created dummy variables for three reform periods: pre-1975 (pre-reform), 1975–1989 (middle reform), and 1990–1996 (modern reform).[154] They created a regression model that included the reform variable as an independent variable (among other variables) and reported rapes as the dependent variable.[155] In the model, rapes were 88% more likely to be reported in the modern period than the pre-reform period.[156] Rapes were not more likely to be reported in the middle reform period than the pre-reform period.[157]

Controlling for Other Factors

Researchers who have used control groups or time-series analysis have not found strong evidence that the legal interventions had an impact on case processing. As compared to control groups, researchers have found no processing change, no unique increases, or only these modest findings: an increase in the percentage of victims who reported, a slight upward trend in the percentage of arrests for rape that resulted in a case filing, and an increase in the probability that an offender would be incarcerated.

Marsh and colleagues looked at rape reporting and processing statistics in Michigan from 1972–1978.[158] Using time-series analysis that compared trends in rape processing with those in murder, aggravated assault, and robbery, they found no relationship between an upward reporting trend and the change in the law.[159] They reported that the improving arrest rate for forcible rape as compared to other crimes was influenced by the change in the law, but only at p < .10, and conviction rates were influenced, but only at
p < .06 .[160] This indicates that the law parameter could be dropped from the model without substantially affecting the model fit.[161] This suggests that the abrupt change in law does not affect case processing, or if it does, not in a way that can be measured.[162]

Schissel looked at the 1983 reforms to the Canadian system using nationwide data from 1969–1990.[163] He conducted time-series analysis and used control groups. The models showed that the rates of arrests and cleared by charging for sexual assault increased dramatically after the change—but so did the rates for non-sexual assault, and the sexual assault cases increased at a lower rate. He concluded that the legislation was not solely responsible for the change; rather, the election of a conservative, law-and-order government at the same time as the reforms caused a change in the processing of all crimes.[164]

Roberts and Gebotys also looked at the Canadian reforms using nationwide data for 1979–1988.[165] Using time-series analysis and control groups, they reported that the post-reform period was associated with significantly more reports of sexual assault and that the rate of increase was twice the rate of increase for non-sexual assault reports.[166] They also reported that police found allegations of sexual assault to be unsubstantiated at nearly the same rate as allegations of non-sexual assault, and the change in rates for sexual assaults “cleared by charge” was matched by similar changes in non-sexual assault cases.[167]

Polk looked at statewide data from California for 1975–1982.[168] During the 1970s, California implemented several reforms (substantive, evidentiary, and penalty) at various times, so this study did not have a precise before-and-after date.[169] Polk used a control group of homicide, arrest, robbery, and burglary. Reporting descriptive statistics, he found that the clearance rate for sexual assaults remained stable, and that pattern held for the control group.[170] The percentage of arrests for rape that resulted in a case filing trended slightly upward with no comparable increase in the control group.[171] The percentage of felony convictions per felony complaint was trendless across all groups.[172] He found a strong upward trend in sentences resulting in prison, but that trend was also found in homicide and burglary, with moderate upward trends in robbery and assault.[173]

Horney and Spohn looked at data from multiple jurisdictions (Detroit, Chicago, Philadelphia, Atlanta, Houston, and Washington, D.C.) that implemented varying degrees of reform on various dates.[174] Using time-series analysis, they looked at data from 1970–1984, ensuring there were at least some years before and after each reform in each jurisdiction.[175] They found that in Michigan, the change in law appeared to have a positive effect on reporting rates, sentencing, and the ratio of indictments to reported cases, but not on the ratio of convictions to indictments.[176] They reported that the rape reforms did not appear to have any effect in the other jurisdictions.[177]

Bachman and Paternoster used national data from the United States from 1973–1990 to detect reporting trends in rape, robbery, and assault.[178] Using survey data, they found that the proportion of rape victims who reported their rapes to police increased by 10% following legal reforms, while the proportions of reported assaults trended upward but to a lesser degree, and the proportion of reported robberies trended downward.[179] Using different data from the Uniform Crime Reports,[180] they found that per capita reports of rape to police trended upward for rape by 13%, while robbery trended upward by 6% and assault by 46%.[181] Further, they found that, since 1981, the probability that an offender arrested for rape would go to prison increased by over 200%, for robbery increased by 9%, and for assault by 25%.[182] Their study was a long-term study that did not include abrupt interventions, and so it likely measured true changes in norms over that period.[183]

Frank and colleagues measured fifty-one reforms in forty countries from 1945 to 2005.[184] They used countries that had police data available for three to five years before and after the reform.[185] They found a 75% increase in reporting during reform periods but only a 3% increase during non-reform periods.[186] They did not use other crimes as a control group but did compare countries that had made reforms to countries that had not made reforms and found that reporting increased even in countries that had not implemented reforms.[187]

Finally, Carpenter and colleagues analyzed the 2007 reform to the military’s rape statute.[188] This reform changed the statute from a common law scheme to a hybrid reform model, roughly based on the Michigan statute.[189] They analyzed all reported sexual assaults in the Army from 2004 to July of 2012 and used all reported non-sexual assaults for the same period as a control group.[190] They found that the ratio of founded sexual assaults to founded non-sexual assaults increased after the change, but when they ran a time-series analysis on the data, the models did not show that the legal intervention had a statistically significant effect on case processing.[191] This suggests that other long-term trends accounted for the increase rather than the change in the law.

In sum, the current research does not provide compelling evidence that legal interventions affect case processing downstream of the report. Researchers looking at only sexual assault have found some limited changes in processing following reforms (such as increases in victim reporting, cases bound over for trial, the ratio of indictments to reported cases, and conviction rates), but most of those studies did not control for outside factors. Researchers who have included a control group or who have used time-series analysis (or both) have provided modest findings: an increase in victim reporting, a slight upward trend in the percentage of arrests for rape that resulted in a case filing, and an increase in the probability that an offender would be incarcerated. There is little evidence about what might cause those unique processing changes—the legal reforms, or changing norms, or some combination of both.

Normative Words and the Fundamental Limits of Rape Law Reform

Looking at this body of research, many have concluded that legal reforms have been ineffective, and that is because legal institutions are resistant to reforms. The laws change, but the attitudes of those who execute the laws do not.[192] Mimi Ajzenstadt and Odeda Steinberg note that the effectiveness of any legal reform depends on how that law is enforced and further explain that there is considerable room within the law for a law enforcement official’s values and norms to enter. [193] If those values and norms remain unchanged, then the formal legal change will have no effect.

More precisely, certain words in the law can only be resolved by using values and norms. These are normative words, and they include: reasonable, should, fair, due, called for by the circumstances, gross deviation in the standard of care, unjustifiable, sufficient, necessary, foreseeable, and offensive. The law does not define these terms in meaningful ways. Instead, law enforcement officials, prosecutors, judges, and jurors are expected to decide when these words are satisfied by using their life experiences and values. If legal actors use norms and values that are racist, homophobic, xenophobic, or sexist, then the normative words fail us. These words become entry points for bias. In the sexual assault context, if legal actors use inaccurate beliefs about rape to decide when these words are satisfied, then these normative words become the gateways for that bias.

The differences between the reform models and the common law are modest to start, and the key is that normative words that were in the common law before the reforms are still in the law after the reforms. One common reform to the substantive law did eliminate a normative word. The substantive common law contained a normative word in the definition of force (reasonable resistance).[194] The reform models have generally dropped that definition.[195] However, there are two places where normative words still exist in the reform models and allow legal actors to solve these cases using rape myths. One of these places is the element of consent and the other is the mistake of fact defense.[196] While both can be tweaked around the edges, they are fundamentally fixed and represent the fundamental limits on rape law reform.

The Consent Element and Victim Credibility

As discussed above, the victim’s credibility as it relates to consent is an entry point for rape myths, and lack of consent is an element of a sexual assault in all of the models.[197] In the force-centric model, it is still part of the crime.[198] It is just not a written element that the government must prove in its case-in-chief. Instead, the defense raises it.[199] Then, the government must still prove a lack of consent beyond a reasonable doubt.[200] The only difference between the models is who has the burden of production on the issue.[201] Once consent is ultimately raised, the credibility of the victim is still on trial regardless of the model—and jury instructions commonly tell jurors to evaluate the reasonableness of the witness’s testimony when deciding whether they are telling the truth, using language like this:

You bring to this process all of your varied experiences. In life, you frequently decide the truthfulness and accuracy of statements made to you by other people. The same factors used to make those decisions, should be used in this case when evaluating the
testimony . . . . Was the testimony of the witness plausible and likely to be true, or was it implausible and not likely to be true?[202]

Those instructions reflect how people decide if someone is telling the truth throughout all aspects of social life. If what someone says is reasonable, it is more likely to be believed. If it is unreasonable, it is less likely to be believed. In trials, when deciding whether someone is telling the truth, legal actors do the same thing. They generate narratives about how people behave so they can spot where the witness is testifying consistently with those expectations (and so telling the truth) or inconsistently (and so is mistaken or telling a lie).[203] If legal actors use rape myths to form these narratives, then they will make inaccurate credibility assessments. The defense will look for inconsistencies with these narratives (“You say you didn’t consent, but you did this? And this? And this? Oh, really.”) and these lines of inquiry are often victim-blaming.

Legal actors also look for reasons why the witness may be lying.[204] When the defense counsel argues that the victim consented, the defense counsel also needs to explain why she is now saying she did not. Showing that she had a motive to lie is one way to do that (the defense can also say she was mistaken). From the defense perspective, these motives could be that she is vindictive, covering up for an affair, regrets drunken sex, or some other victim-blaming motive. If these legal actors believe that women frequently make false allegations for these reasons, then they will be more likely to believe this victim lied on this occasion

Rape shield rules were supposed to limit the introduction of victim-blaming evidence, but when those exclusionary rules interfere with the defense’s ability to cross-examine the victim, the Confrontation Clause is implicated.[205] The Confrontation Clause is also normative. Trial judges can only impose “reasonable limits on such cross-examination.”[206] If a trial judge does impose limits, those limits must still allow defendants to engage in “otherwise appropriate cross-examination designed to show a prototypical form of bias on the part of the witness, and thereby ‘to expose the jury the facts from which jurors could appropriately draw inferences relating to the reliability of the witness.’”[207]

Many courts find that appropriate or reasonable lines of cross-examination[208] include those where the victim’s prior consensual sex acts are sufficiently similar to the current sex acts (especially when the acts are unusual), with the inference being that if she consented to those sex acts with others in the past (she is predisposed to having this kind of sex with other people), she was more likely to have consented with the defendant on the occasion in question.[209] The rape shield rule in the Federal Rules of Evidence includes a similar exception, such that evidence that the victim had sex with the defendant on a prior occasion is admissible to prove consent.[210] The inference to draw is that, if she consented to sex with him before (she was predisposed to have sex with him in particular), then she is more likely to have consented on this occasion.

Whether these patterns are sufficiently probative of consent and whether evidence of bias is necessary under the Confrontation Clause is left up to judges, and if the judge, because of his or her values and life experiences, believes women who acted like that must have consented, or that women routinely lie about rape, then the judge may allow the evidence to come in. The jury may then process the problem the same way.

This is a fundamental limit on rape law reform. Legislatures cannot get rid of the element of consent and the focus on victim credibility. All legislatures can do is delay their appearance at trial. The Confrontation Clause governs the cross-examination of the complaining witness, and the Confrontation Clause is itself normative and fixed. No measure of legal reform other than amending the Constitution can close this entry point for rape myths.

The Mistake of Fact Defense

As discussed above, the mistake of fact defense is an entry point for rape myths, and the reform models retain the mistake of fact defense.[211] Both before and after the legal intervention, if the victim did not subjectively consent, legal actors must then consider whether the offender reasonably believed that the victim had consented based on her manifestations of consent that the offender observed.[212] If the offender appears ingroup, and the fact pattern otherwise looks like typical consensual sex scenarios, the legal actors need to resolve the dissonance between the victim’s allegation and the rule that ingroup men do not rape. One way to do that is to conclude that this must have been a mistake. The evidence that supports this defense often focuses on the victim and is usually victim-blaming. It is evidence of the victim’s sexual behaviors before the assault that might have led the offender to believe that the victim was predisposed to have sex with him and then acted in conformity with that predisposition at the moment of the sex act.

Legislatures passed rape shield laws to prevent the admission of victim-blaming evidence, but when those exclusionary rules interfere with the defense’s presentation of the mistake of fact defense, the Due Process Clause is implicated.[213] The Due Process Clause is fundamentally normative. The normative word due is further described using other normative words, like fair: “The right of an accused in a criminal trial to due process is, in essence, the right to a fair opportunity to defend against the State’s accusations.”[214] What it means to be fair is further described by other normative words such as “a meaningful opportunity to present a complete defense,”[215]essential to his defence,”[216] and “necessary . . . to present [the] defense.”[217]

If the judge, because of his or her values and life experiences, believes that an ingroup man might misunderstand the situation because the victim had sexual relationships with other men and the ingroup man knew about it, or the victim wore certain clothes, or she danced with him a certain way, or she did not fight back, then the judge may find that that evidence is central (or essential or necessary) for the defendant to have a meaningful opportunity to defend himself.[218] When the evidence is introduced, jurors may process that evidence the same way to resolve the dissonance between the allegation and the generalization that ingroup men do not rape. The substantive law changed, and rape shield rules arrived, but the mistake of fact defense still exists, and rape myths are thus able to enter the system.

Technically, the mistake of fact defense is not fundamentally fixed because it is not constitutionally required. The mistake of fact defense is a burden of proof defense related to the offender’s mental state, and it is constitutional to have crimes that do not include a mental state.[219] Moreover, in a statute that reads, “without the other person’s consent, does knowingly commit a sex act,” there is a mental state—knowingly (or voluntarily) committing a sex act.[220] In that statute, there is no requirement that the offender know that the attendant circumstance of “consent” exists. That is true in statutory rape cases, where knowledge of the attendant circumstance of the victim’s age (or a mistake related to the existence of that attendant circumstance) is often not required.[221] The Supreme Court, in the context of interpreting a federal statute and relying on interpretive principles rather than constitutional mandates, has said that, when Congress has not been clear that an element does not have a mental state, then the Court will infer a mental state that will separate wrongful conduct from otherwise innocent conduct.[222] In a sexual assault case, that would likely mean inferring negligence as to the existence of an attendant circumstance, which is the current state of the law with the mistake of fact defense.[223]

Importantly, the Supreme Court will not infer a mental state when Congress is clear that no mental state attaches.[224] If Congress (or another legislative body) is clear that a mental state does not attach to the attendant circumstance of consent (as legislatures are with the attendant circumstance of age in statutory rape), then not having a mental state would be constitutional. And if the defendant does not have to know that the attendant circumstance exists, then it does not matter if he was mistaken. The mistake of fact defense is irrelevant, constitutionally so.

Whether to have a mistake of fact defense related to sexual assault or not is a matter of public policy, not a constitutional mandate. The better public policy would be to shift the risk about whether consent exists from the person on whom the sex act is being committed to the person committing the sex act. And it would help close the door on evidence related to a primary rape myth—that ingroup men do not rape, and so this must have been a mistake. That said, it is very unlikely that any legislative body will eliminate it, although there is room for reform as the next Part suggests. However, because the mistake of fact defense could be eliminated, it is not technically a fundamental limit, but it still uniformly exists and acts as a fundamental limit. It will continue to serve as an entry point for rape myths.

The Best We Can Do?

Consent (and therefore victim credibility) is still central in all reform models, and jurisdictions have universally retained the mistake of fact defense. Both the consent element and the mistake of fact defense are conduits for rape myths. They both still exist, and they both allow rape myths to enter the system. Therefore, we should not expect legal changes to have much effect on rape case processing, which is something that the research reflects.[225] The substantive law may have changed, and rape shield rules arrived, but those laws and rules are themselves governed by the Bill of Rights, and that set of laws is fundamentally normative and fundamentally fixed.

Faced with these fundamental limits, some reforms may still be worthwhile. Each of these proposals could generate a separate article, and the goal of this Article is just to briefly discuss some potential reforms from the perspective of limiting the effect that rape myths will have as a result of the continued presence of normative words in the law. Abstract normative words will always be there, but legislatures can list examples of concrete factual situations and then limit the legal actors’ discretion when they evaluate those factual situations. This will remove these situations from the normative space that normative words create.

Define the Legal Meaning of Communications

If the law cannot sort out the legal meaning of certain communications, we cannot expect the legal actors to be able to—and we should then expect that they will rely on “mythcommunication” beliefs to solve the problem in front of them. First, we need to focus on how legal actors use these communications. The victim will likely give direct evidence that she did not consent by giving a statement to the police, or later, by testifying in court. Legal actors then look to her outward manifestations of consent to evaluate whether she is telling the truth and also to evaluate whether the offender had a reasonable belief that she consented. The law needs to tell legal actors what to do with manifestations of yes, what to do with silence, and what to do with manifestations of no.[226]

Starting with yes, the Canadian statute reflects a common definition of consent: consent is a freely given agreement to sexual conduct.[227] For direct evidence, the victim would say, “I did not agree to have that sexual contact.” The tougher problem for the legal actors is sorting through the other circumstances in the case that contradict her statement that she did not agree (the circumstances that suggest she said yes) and which might also support the offender’s claim that he was reasonably mistaken about consent because of existence of those circumstances.

Statutes should make clear that, in sexual situations, people do not need lawyers in the bedroom drafting contracts for consent.[228] Just like in many areas of life, a yes or freely given agreement can be nonverbal.[229] Toward that, many statutes say consent may be determined by looking at the totality of the circumstances. For example, the military’s statute says that all of the surrounding circumstances can be considered in determining whether a person consented.[230]

However, statutes should do better. When impeaching the victim’s statements or deciding whether the offender was reasonably mistaken, all of the surrounding circumstances should not be considered. Just some of them should be.[231] Statutes should narrow the period that matters to the moments just before the sex act. This will focus the legal actors on whether the victim consented or the offender was reasonably mistaken at that moment. The Canadian statute contains a line that gets close to doing this: “Consent must be present at the time the sexual activity in question takes place.”[232] The victim’s behaviors leading up to the moment of the sex act may indicate that the victim was predisposed to consent in the future (in an hour, or later that night, or a couple of weeks later, or a year later), but they do not show that the victim consented at the moment.[233] A predisposition to consent just means someone might consent to something in the future. However, an individual’s behaviors at the moment show whether that person actually consented.

If the victim gave positive signals earlier, that is not relevant to the communications she gave in the moments immediately before the sex act, and that is the moment the law should care about. Someone can say fifty times, “I think I want to buy that house in the future,” but that does not inform what happened at the moment of signing. What should matter to the law is what happened at the time of closing. Did the person say in their lawyer’s office with the papers on the desk in front of them, “I will buy this house?” In a dispute, the evidence that matters is what happened in those immediate moments. And when looking at whether the offender could have been reasonably mistaken because of these earlier circumstances, any evidence of earlier, positive signals only shows that at the moment of the sex act, he was mistaken about her predisposition, not whether she was consenting at that moment. We will return to this thread in the next Section.

In addition to narrowing the period for the circumstances that may be used to impeach the victim or to support a reasonable mistake of fact, statutes should also define what circumstances may not be considered. For example, the military’s statute contains this line: “A current or previous dating or social or sexual relationship by itself, or the manner of dress of the person involved with the accused in the conduct at issue, does not constitute consent.”[234] If the circumstances that play into rape myths are specifically excepted from the circumstances that may be considered for impeachment or to support a reasonable mistake, then evidence of those victim behaviors becomes irrelevant to that crime as defined by the legislature.

Turning to silence, in consent-centric statutes, a sex act without consent is unlawful. Before someone can perform a sex act on someone else, the other person must take some affirmative step through words or actions to indicate consent.[235] Therefore, silence without other action means the same thing as no. Statutes should say that explicitly. Force-centric models need to anticipate that the defense will raise the issue of consent (to say there was no force) and set the presumption of consent as “no, unless,” such that silence cannot equal consent. When the law states that silence equals “no,” legal actors should recognize the intuitive truth behind that presumption. Someone cannot take your property without first hearing you say “yes” through words or actions. Likewise, someone cannot touch your body in certain places unless they first hear you say (through words or actions) “yes.”

Statutes should not do what the American Law Institute did in its recently approved definition of consent, which is to allow the victim’s inaction to be a factor that may be considered.[236] The recently-drafted comment to Section 213.0(2) states: “[I]f a sexual act is clearly foreshadowed and nothing suggests an impediment to the other person’s ability to object to it, the totality of that person’s conduct, including both acts and omissions, can be considered in determining whether that person consented.”[237] This language invites the legal actor to use predisposition evidence (“clearly foreshadowed”) and then allows inaction at the moment of the sex act to be enough. Under that language, predisposition plus inaction equals evidence of consent. Using this reasoning, if someone says fifty times, “I think I want to buy that house,” and then at the moment of closing stays silent, then that silence can help prove that the person consented to buying the house. Instead, statutes should state that silence, inaction, or lack of resistance are not factors that can be considered when impeaching the victim about her consent at that moment or when determining whether the offender made a reasonable mistake.

Turning to no, the Canadian statute says that no is any expression of no by words or conduct, before or even during an ongoing consensual sex act.[238] The military’s statute defines it much the same way: “An expression of lack of consent through words or conduct means there is no consent.”[239] Both recognize that a “soft no” counts.[240] In both, if the victim manifests those words or actions, then as a matter of law, there is no consent. The defense can still argue that the victim is not telling the truth about what she said and did at that moment, but making this clear helps to close the door on the “mythcommunication” and makes an argument that the offender was reasonably mistaken much less plausible. This change reflects that using a “soft no” is ubiquitous in society, and men (and women) routinely recognize when they are getting a “soft no” and then honor that “no.”

Reform the Mistake of Fact Defense.

While it is very unlikely that any jurisdiction will eliminate the mistake of fact defense, legislatures can define those situations where the offender’s mistake is not reasonable, thereby taking those situations out of the legal actor’s normative evaluation. The Canadian version is a good example.[241] Under this version of the mistake of fact defense, the defense is unavailable if the accused’s belief is due to his intoxication, reckless conduct, or willful blindness; or if he did not take reasonable steps to determine if the victim was consenting; or when “there is no evidence that the complainant’s voluntary agreement to the activity was affirmatively expressed by words or actively expressed by conduct.”[242]

This reform addresses the generalization that ingroup men do not rape by showing that these behaviors are indeed deviant and blameworthy, and this reform shifts the focus to the offender’s culpable behavior (recklessness or willful blindness) rather than the victim’s behavior. Further, this reform chips away at the “mythcommuniciation” by stating explicitly what we already know: when faced with a partner who is sobbing, turning cold, or otherwise showing that she is not interested in sexual behavior, normal men take reasonable steps to see what is wrong (“Huh? I thought things were going well.”). Last, this reform shifts more of the risk about whether there is consent to the person who is trying to commit the sex act. Now, rather than putting the legal burden on the person who does not want sex to clearly communicate the lack of consent, the law puts the burden on the person seeking to have sex to clarify the situation.

These changes, along with reformed language related to the legal meaning of communications, should also reduce the relevance of the victim’s earlier behaviors to the mistake of fact defense. If an offender was getting positive signals earlier in the evening, and now at the moment that consent counts—the moments just before the sex act—is getting opposite signals, he is on additional notice that he needs to take reasonable steps to clear up that change in signals. If someone needs to take reasonable steps to determine consent, that is especially true when that person earlier received positive signals and is now receiving negative signals. This evidence now favors the prosecution. Because evidence of the victim’s behaviors has lowered relevance for the mistake of fact defense, a judge can more easily exclude it as not being necessary for that defense. The necessary evidence to determine if he was mistaken at that moment is the evidence surrounding that moment, not the evidence from several hours earlier.

Clean Up Rape Shield Rules

As discussed above, rape shield rules have limited value because they are governed by the Bill of Rights, and those rules are fixed and fundamentally normative.[243] Essentially, rape shield rules do two things. They increase the relevance threshold for this type of evidence from a simple relevance standard (any tendency to make consent more or less probable or any tendency to make the mistake of fact defense more or less probable)[244] to a very high relevance standard for substantive evidence (the evidence must be central or essential or necessary to the defense) and a higher relevance standard for bias (the inquiry must be appropriate).[245] And, because of notice provisions, the rules force the parties to slow down and fully articulate the reason for or against admission.[246] The rules do not provide a complete barrier to this information. They are a speed bump, not a shield.

Some provisions in rape shield rules make things worse. Some incorporate rape myths directly. As we saw, the Federal Rules of Evidence include an exception for previous instances of sex between the victim and the defendant if offered to prove consent.[247] The inference to draw is, that if she consented to sex with him before, then she was predisposed to have sex with him now. That endorses the rape myth that this kind of rape (within a relationship or marriage) is not real rape—consent is presumed in a relationship, or at least, it should be easier to prove. Jurisdictions that have this exception need to remove it. The focus should be on whether the victim consented to this sex act, not to previous sex acts.

Further, many rape shield statutes contain an explicit constitutional exception to the statute, meaning evidence of the victim’s sexual predisposition and specific acts of sexual behavior are allowed if their exclusion would violate the Constitution.[248] Here is the exception found in the Federal Rules of Evidence: “The court may admit the following evidence in a criminal case: evidence whose exclusion would violate the defendant’s constitutional rights.”[249] Those exceptions just state a truism. No rule of evidence can prohibit the admission of constitutionally required evidence.

Yet no other rule in the Federal Rules of Evidence, to include exclusionary rules, includes this language. For example, another rule that can exclude evidence based on its probative value is Federal Rule of Evidence 403.[250] Among other things, that rule can exclude evidence whose probative value is substantially outweighed by unfair prejudice. But the drafters did not include, “unless constitutionally required” into the rule—because it is not needed. The Constitution is always there in the background. If the application of the rule violates the Constitution, then the defense can make a motion and argue why that evidence is so extraordinarily probative that no level of danger can exclude it.

However, notice what happens when “unless constitutionally required” is included in the rape shield rule. Including this language works to create a presumption that the evidence is normally constitutionally required. For this one type of crime, we stop and tell judges, “Be on alert! This evidence could be constitutionally required! See? Look at this warning!” Judges routinely exclude evidence under Federal Rule of Evidence 403, and that rule does not tell judges, “Be careful! The evidence could be constitutionally required!” Yet the rules do it here, and this feeds into the idea that somehow the defendants in rape cases are exceptional. The rule reflects that these offenders are from the ingroup, and so judges need to take care that their rights are not violated. Federal Rule of Evidence 412(b)(1)(B) communicates that message much in the same way that Army leadership told its law enforcement officers to take special care in sexual assault cases with those particular offenders.[251]

Legislatures should remove this language from rape shield rules. If the defense believes that the exclusion of the evidence would violate the Constitution, the defense can file a motion and argue why the evidence is so extraordinarily probative that the defense cannot get a fair trial without it.


This Article ultimately argues that we should not expect legal reforms to impact case processing in sexual assault cases because of the normative words that exist in the law. These words are the entry points for rape myths. They are in the pre-reform laws, they are in the post-reform laws, and they will always be there. They exist in the evaluation of victim credibility, which is governed by the Confrontation Clause, and which is fundamentally normative and fundamentally fixed. Normative words exist in the mistake of fact defense, which legislatures are very unlikely to eliminate. This defense is governed by the Due Process Clause, which is also fundamentally normative and fundamentally fixed.

While some reforms may reduce the entry points for rape myths, they can only do so much. Even with reformed language related to proving consent, judges may still think that the victim’s behaviors from earlier in the evening are necessary for the defendant to argue that he reasonably believed that she had consented. Even with reformed language related to consent that focuses the legal actors on the moments and circumstances that matter, the defense can always say that the victim is lying about the circumstances that happened in those moments, and impeachment of the victim (and other extrinsic evidence of bias) will serve as entry points for inaccurate generalizations about rape. Even under the Canadian version of the mistake of fact defense, legal actors could rely on inaccurate generalizations about rape to decide that the offender took reasonable steps to learn if the victim was consenting. The entry points are still there.

The normative words in the law are not going to change. That is the fundamental limit of rape law reform. The only thing that can change is the norms that we apply to them. That could happen at the macro level, with norms changing at the societal level—but that takes a long time. This means that legal actors need to change norms at the micro level, within the investigative and trial phases of a particular case. If the words are going to be there, we need to change a legal actor’s belief systems—maybe not for good, and maybe just for this case—before a legal actor processes these normative words.

Law enforcement officers and prosecutors need to be individually trained on offender and victim behaviors. They need to be told, contrary to what the Army told its investigators, that they should have an investigative presumption that the offender committed the alleged offense and that the victim is telling the truth.[252] They should conduct investigations to find evidence that shows that the offender engaged in blameworthy behavior and to find evidence that is consistent with the victim’s story (all while remaining alert to red flags of false allegations). They need to be trained in modern methods for interviewing sex assault victims that do not treat the victim as the subject of cross-examination.[253]

We can ensure that these legal actors use accurate rape generalizations when processing cases, and if they do, we should expect a gradual improvement in case processing. We cannot control the belief systems of those who report for jury duty, so prosecutors need to be trained to tell stories at trial that highlight the offender’s premeditated and intentional behavior[254] and that explain victim behaviors that do not fit the behaviors of rape victims that many people expect, and so are counterintuitive.[255] By showing that, in a given case, the offender’s behaviors were deviant and the victim’s behaviors were quite normal, the jury can then see that this was indeed a real rape. As these successful prosecutions begin to aggregate over time, we may begin to see the positive trends in case processing that we had hoped would be brought about by reforming the law. Through careful investigative and prosecutorial practices that protect victims and tell their stories deliberately and truthfully, we can move past the fundamental limits of rape law reform.

  1. * Associate Professor of Law, Florida International University College of Law. The author thanks Tracy Pearl, Joelle Moreno, and Howard Wasserman for comments provided during the drafting process.

  2. . Kimberly A. Lonsway & Joanne Archambault, The “Justice Gap” for Sexual Assault Cases: Future Directions for Research and Reform, 18 Violence Against Women 145, 157 (2012).

  3. . See David P. Bryden & Sonja Lengnick, Rape in the Criminal Justice System, 87 J. Crim. L. & Criminology 1194, 1201–11 (1997); Martha R. Burt, Rape Myths and Acquaintance Rape, in Acquaintance Rape: The Hidden Crime 26, 27 (Andrea Parrot & Laurie Bechhofer eds., 1991).

  4. . See Susan Estrich, Rape, 95 Yale L. J. 1087, 1094–1134 (1986).

  5. . See id. at 1157–61; see also infra Part III (discussing the effects of rape law reforms).

  6. . See Estrich, supra note 3, at 1099.

  7. . See infra Part IV.

  8. . See infra Section IV.A. I recognize that the Constitution can be amended, but am confident that the Due Process Clause and Confrontation Clause will not be amended anytime soon.

  9. . See infra Section IV.B.

  10. . Bryden & Lengnick, supra note 2, at 1207. In this context, consent-defense cases are assaults by men of women where the defendant can plausibly raise the defense of consent. See id. at 1204; see also Eric R. Carpenter, The Military’s Sexual Assault Blind Spot, 21 Wash. & Lee J. Civ. Rts. & Soc. Just. 383, 389 n.27 (2015) (“Recognizing that men are also victims of sexual assaults and that women can also commit sexual assault, this article is focused on the sexual assaults that are the focus on feminist critical theory: sexual assaults of adult women by adult men.”).

  11. . See Lonsway, supra note 1, at 159; Burt supra note 2, at 33.

  12. . Kimberly A. Lonsway & Louise F. Fitzgerald, Rape Myths: In Review, 18 Psych. Women Q. 133, 134 (1994).

  13. . Diana L. Payne et al., Rape Myth Acceptance: Exploration of Its Structure and Its Measurement Using the Illinois Rape Myth Acceptance Scale, 33 J. Res. Personality 27, 42 (1999); see Burt supra note 2, at 28–33.

  14. . I discuss this theory and the social and cognitive psychology that underlies it at greater length in another article. See Carpenter supra note 9, at 390–07.

  15. . See Burt supra note 2, at 27.

  16. . Carpenter, supra note 9, at 397.

  17. . Id.

  18. . See Burt supra note 2, at 27; Carpenter, supra note 9, at 397.

  19. . See David Lisak & Paul M. Miller, Repeat Rape and Multiple Offending Among Undetected Rapists, 17 Violence & Victims 73, 78 (2002); Kevin M. Swartout et al., Trajectory Analysis of the Campus Serial Rapist Assumption, 169 JAMA Pediatrics 1148, 1150 (2015); Antonia Abbey et al., Attitudinal, Experiential, and Situational Predictors of Sexual Assault Perpetration, 16 J. Interpersonal Violence 784, 793 (2001); see also Michele L. Ybarra & Kimberly J. Mitchell, Prevalence Rates of Male and Female Sexual Violence Perpetrators in a National Sample of Adolescents, 167 JAMA Pediatrics 1125, 1129 (2013) (finding similar statistics for reports of sexual violence perpetrated by adolescents).

  20. . Brad J. Bushman et al., Narcissism, Sexual Refusal, and Sexual Aggression: Testing a Narcissistic Reactance Model of Sexual Coercion, 84 J. Personality & Soc. Psych. 1027, 1037–38 (2003); see Jeffrey A. Bernat et al., Sexually Aggressive and Nonaggressive Men: Sexual Arousal and Judgments in Response to Acquaintance Rape and Consensual Analogues, 108 J. Abnormal Psych. 662, 671 (1999); Linda J. Skinner & Kenneth K. Berry, The Perpetrators of Date Rape: Assessment and Treatment Issues, in Acquaintance Rape: Assessment, Treatment, & Prevention 55, 56–57 (Thomas L. Jackson ed., 1996).

  21. . David Lisak, Understanding the Predatory Nature of Sexual Violence, 14 Sexual Assault Rep. 49, 56 (2011); Veronique N. Valliere, Understanding the Non-Stranger Rapist, 1 Voice 1, 2 (2007).

  22. . Mary P. Koss et al., Stranger and Acquaintance Rape: Are There Differences in the Victim’s Experience?, 12 Psych. Women Q. 1, 12 (1988).

  23. . See Bernat, supra note 19, at 670.

  24. . Valliere, supra note 20, at 4.

  25. . See Teresa P. Scalzo, Am. Prosecutors Rsch. Inst., Prosecuting Alcohol-Facilitated Sexual Assault 27 (2007).

  26. . See Burt supra note 2, at 28.

  27. . See id.

  28. . See id. at 31–32.

  29. . See Carpenter, supra note 9, at 399; Scalzo, supra note 24, at 27; Patricia L. Fanflik, Am. Prosecutors Rsch. Inst., Victim Responses to Sexual Assault: Counterintuitive or Simply Adaptive? 20 (2007).

  30. . Carpenter, supra note 9, at 400.

  31. . See id. at 399.

  32. . See Catharine A. MacKinnon, Marxism, Method, and the State: Toward Feminist Jurisprudence, 8 Signs 635, 654 (1983).

  33. . See Carpenter, supra note 9, at 398.

  34. . Jodee M. McCaw & Charlene Y. Senn, Perception of Cues in Conflictual Dating Situations, 4 Violence Against Women 609, 610 (1998).

  35. . Melanie A. Beres et al., Navigating Ambivalence: How Heterosexual Young Adults Make Sense of Desire Differences, 51 J. Sex Rsch. 765, 773 (2014), Some of this miscommunication hypothesis flows from a strand of research that found that upwards of 40% of women engaged in token resistance behaviors. See Charlene L. Muehlenhard & Lisa C. Hollabaugh, Do Women Sometimes Say No When They Mean Yes? The Prevalence and Correlates of Women’s Token Resistance to Sex, 54 J. Personality & Soc. Psych. 872, 874 (1988). This research was influential. For example, Stephen Schulhofer cited it before concluding that, “[M]istakes—including reasonable mistakes—are not impossible or even rare. Sexual miscommunication is so often indirect and contradictory that it is a wonder mistakes do not occur more often.” Stephen J. Schulhofer, Unwanted Sex 65 (1998); see id. at 259–60. However, the principal author of that study later recognized that the instrument she used (and which many subsequent researchers have used) is invalid and unreliable. She found that, “Respondents’ narratives often indicated that they had misinterpreted our questions.” See Charlene L. Muehlenhard & Carie S. Rodgers, Token Resistance to Sex: New Perspectives on an Old Stereotype, 22 Psych. Women Q. 443, 448–49 (1998). And she concluded that, “Consequently, it seems likely that the percentages reported in previous studies overestimated the actual prevalence of token resistance.” See id. at 461. See also Charlene L. Muehlenhard, Examining Stereotypes About Token Resistance to Sex, 35 Psych. Women Q. 676, 678 (2011).

  36. . Thomas MacAulay Miller, Mythcommunication: It’s Not That They Don’t Understand, They Just Don’t Like the Answer, Yes Means Yes (Mar. 21, 2011), [].

  37. . See Melanie Beres, Sexual Miscommunication? Untangling Assumptions About Sexual Communication Between Casual Sex Partners, 12 Culture, Health & Sexuality 1, 11 (2010); Celia Ketzinger & Hannah Frith, Just Say No? The Use of Conversation Analysis in Developing a Feminist Perspective on Sexual Refusal, 10 Discourse & Soc. 293, 309 (1999); Jodee M. McCaw & Charlene Y. Senn, Perception of Cues in Conflictual Dating Situations, 4 Violence Against Women 609, 621–22 (1998); Rachael O’Byrne et al., ‘You Couldn’t Say “No”, Could You?’: Young Men’s Understandings of Sexual Refusal, 16 Feminism & Psych. 133, 149 (2006).

  38. . Beres, supra note 36, at 7–8; see also Ketzinger & Frith, supra note 36, at 309–11 (discussing various methods that women utilize when responding to unwanted sexual pressure).

  39. . See Beres et al., supra note 34, at 9 (explaining that most men correctly identify a woman’s refusal to engage in sexual intercourse); Rachael O’Byrne et al., “If a Girl Doesn’t Say ‘No’ . . .”: Young Men, Rape and Claims of ‘Insufficient Knowledge’, 18 J. Cmty & Applied Soc. Psych. 168, 187 (2008); see also O’Byrne et al., supra note 36, at 149; McCaw & Senn, supra note 33, at 622 (discussing issues associated with the miscommunication hypothesis). See generally Beres, supra note 36, at 5–11 (analyzing the methods with which men are able to identify a potential partner’s sexual communication).

  40. . See McCaw & Senn, supra note 33, at 616 (stating that most men are “keenly aware” when a woman refuses consent in sexual situations).

  41. . See Bushman et al., supra note 19, at 1039 (stating types of situations in which men use force in sexual situations); Lisak, supra note 20, at 4 (analyzing sexual assaults involving entitlement and uses of force). See generally Bernat et al., supra note 19, at 670–72 (discussing sexual assault situations in which aggressive behavior is a factor).

  42. . Miller, supra note 35, at 1.

  43. . Eric R. Carpenter, Differential Treatment of Sexual Assault Cases by US Army Law Enforcement Personnel, Violence Against Women (forthcoming 2022).

  44. . See Donald Johnson et al., Use of Forensic Science in Investigating Crimes of Sexual Violence: Contrasting Its Theoretical Potential With Empirical Realities, 18 Violence Against Women 193, 213 (2012); Cassia Spohn & David Holleran, Prosecuting Sexual Assault: A Comparison of Charging Decisions in Sexual Assault Cases Involving Strangers, Acquaintances, and Intimate Partners, 18 Just. Q. 651, 682 (2001); Wayne A. Kerstetter, Gateway to Justice: Police and Prosecutorial Response to Sexual Assaults Against Women, 81 J. Crim. & Criminology 267, 301, 307 (1990).

  45. . See Steffen Bieneck & Barbara Krahé, Blaming the Victim and Exonerating the Perpetrator in Cases of Rape and Robbery: Is There a Double Standard?, 26 J. Interpersonal Violence 1785, 1795 (2011) (discussing the potential for bias when conducting research regarding rape cases).

  46. . Cassia Spohn et al., Prosecutorial Justifications for Sexual Assault Case Rejection: Guarding the “Gateway to Justice, 48 Soc. Probs. 206, 233 (2001) (discussing the impact of stereotypes in cases involving rape and other forms of sexual assault); Lisa Frohmann, Discrediting Victims’ Allegations of Sexual Assault: Prosecutorial Accounts of Case Rejections, 38 Soc. Probs. 213, 213 (1991) (analyzing the perception of “victim credibility” in sexual assault cases). See generally Amy Dellinger Page, Gateway to Reform? Policy Implications of Police Officers’ Attitudes Toward Rape, 33 Am. J. Crim. Just. 44, 44 (2008) (assessing “rape myths” and associated factors affecting police officers’ decision-making in rape cases).

  47. . See Lisa Frohmann, Convictability and Discordant Locales: Reproducing Race, Class, and Gender Ideologies in Prosecutorial Decisionmaking, 31 Law & Soc’y Rev. 531, 552–53 (1997) (discussing the role of prosecutorial biases in the selection of which sexual assault cases to prosecute); see also Frohmann, supra note 45, at 224 (detailing the connection between rape myths and prosecutorial success in obtaining a conviction).

  48. . See Andrea Quinlan, Suspect Survivors: Police Investigation Practices in Sexual Assault Cases in Ontario, Canada, 26 Women & Crim. Just. 301, 315 (2016) (arguing that “survivors who do not fit the narrow confines of what investigators deem to be normal responses to sexual assault are often viewed with greater suspicion”); Lucy Maddox et al., The Impact of Psychological Consequences of Rape on Rape Case Attrition: The Police Perspective, 27 J. Police & Crim. Psych. 33, 43 (2012) (discussing factors contributing to police officers minimizing the credibility of rape victims when the narrative fails to match preconceived notions of sexual assault). See generally Jan Jordan, Beyond Belief? Police, Rape and Women’s Credibility, 4 Crim. Just. 29, 50–53 (2004) (summarizing the impact associated with police officers “rely[ing] heavily on stereotypes to assist in the general exercise of discretion” with regards to sexual assault investigations).

  49. . See Fanflik, supra note 28, at 19 (discussing issues associated with rape myths).

  50. . See Megan A. Alderden & Sarah E. Ullman, Creating a More Complete and Current Picture: Examining Police and Prosecutor Decision-Making When Processing Sexual Assault Cases, 18 Violence Against Women 525, 541 (2012) (“[C]ases involving discrepancies in victim accounts [are] more likely to result in officers questioning the legitimacy of victim claims.”).

  51. . See Page, supra note 45, at 44 (highlighting the negative prosecutorial consequences associated with rape myths).

  52. . Deborah Tuerkheimer, Incredible Women: Sexual Violence and the Credibility Discount, 166 U. Pa. L. Rev. 1, 33 (2017); see also Bieneck & Krahé, supra note 44, at 1787 (calling this the “leniency bias,” which is to lower perpetrator blame and increase victim blame in rape cases).

  53. . See U.S. Dep’t Army, Army 2020: Generating Health & Discipline in the Force Ahead of the Strategic Reset 121 (2012) (discussing Army leadership efforts to mitigate sex crimes in the military).

  54. . Id. at 129; see David Lisak et al., False Allegations of Sexual Assault: An Analysis of Ten Years of Reported Cases, 16 Violence Against Women 1318, 1319, 1330 (2010) (stating that 2.1–10.9% of rape reports are false).

  55. . See e.g., U.S. Dep’t Army, supra note 52, at 100–02 (detailing various crimes in the military).

  56. . Id. at 129.

  57. . Id.

  58. . See, e.g., id. at 100–01 (discussing trends in serious violent crimes).

  59. . This is not to suggest that law enforcement officers are overtly sexist or intentionally sabotaging these cases. For most, the bias is implicit rather than overt. Because the bias is usually implicit, “these decisions appear rational, necessary, and appropriate” to those who are working the cases. Frohmann, supra note 45, at 214. The legal actors will genuinely think they are doing the right thing when they decide to drop what could otherwise be a meritorious case.

  60. . See Tuerkheimer, supra note 51, at 35 (highlighting problems associated with presumptions by police officers in sexual assault cases).

  61. . See id. at 33–35 (discussing ways in which biases impact investigators during rape cases).

  62. . See id. at 33 (discussing the negative effects of downstream credibility issues in sexual assault situations).

  63. . Rebecca Campbell, What Really Happened? A Validation Study of Rape Survivors’ Help-Seeking Experiences With the Legal and Medical Systems, 20 Violence & Victims 55, 56, 62, 65 (2005).

  64. . See Jody Clay-Warner & Callie Harbin Burt, Rape Reporting After Reforms: Have Times Really Changed?, 11 Violence Against Women 150, 150–51 (2005) (highlighting issues associated with victims not reporting sexual assault situations).

  65. . Susan Caringella, Addressing Rape Reform in Law and Practice 1 (2009).

  66. . See id. at 13 (discussing statutory elements of rape and sexual assault crimes); Schulhofer, supra note 34, at 18.

  67. . Schulhofer, supra note 34, at 18.

  68. . Id.

  69. . Id.; Paul C. Giannelli, Understanding Evidence 155 (5th ed. 2018).

  70. . Schulhofer, supra note 34, at 18.

  71. . See generally Estrich, supra note 3, at 1094–32 (detailing the legal elements associated with common law rape).

  72. . Joshua Dressler, Understanding Criminal Law 559 (9th ed. 2022).

  73. . Id.; see also Joshua Dressler & Stephen P. Garvey, Criminal Law: Cases and Materials 167–68 (8th ed. 2019) (discussing the “elemental” approach to understanding the distinction between general and specific crimes).

  74. . See Dressler, supra note 71, at 66.

  75. . See id. at 66–67.

  76. . Schulhofer, supra note 34, at 31.

  77. . See Estrich, supra note 3, at 1121 n.101 (discussing “nonconsent” as an element of rape).

  78. . See id. at 1126 (mentioning other crimes that lack a specific element of consent).

  79. . Id. at 1121 n.101; see Schulhofer, supra note 34, at 35; Dressler, supra note 71, at 66.

  80. . See Dressler, supra note 71, at 74–75; Estrich, supra note 3, at 1121 n.101; Schulhofer, supra note 34, at 35.

  81. . See, e.g., Estrich, supra note 3, at 1125.

  82. . See id. at 1126–27.

  83. . See Caringella, supra note 64, at 14; Estrich, supra note 3, at 1105–21.

  84. . See Estrich, supra note 3, at 1127–30.

  85. . See Schulhofer, supra note 34, at 258–59; Estrich, supra note 3, at 1097.

  86. . See Dressler, supra note 71, at 152–53.

  87. . See id. at 579–80.

  88. . See id.

  89. . See id.

  90. . See id. at 153 n.10.

  91. . See id. at 580.

  92. . Estrich, supra note 3, at 1095, 1999, 1107.

  93. . Clay-Warner & Burt, supra note 63, at 152.

  94. . See generally Schulhofer, supra note 34, at 17–46; Caringella, supra note 64, at 12–27.

  95. . See Schulhofer, supra note 34, at 30, 43; Caringella, supra note 64, at 14, 16–17, 20.

  96. . Schulhofer, supra note 34, at 30; Caringella, supra note 64, at 15–16; see Giannelli, supra note 68, at 157–60

  97. . Caringella, supra note 64, at 17–18.

  98. . See id. See generally Schulhofer, supra note 34, at 31–33.

  99. . Dressler, supra note 71, at 559.

  100. . See Estrich, supra note 3, at 1133.

  101. . See Mich. Comp. Laws § 750.520b(1)(f) (2017).

  102. . See Criminal Code, R.S.C. 1985, c. C-46, s. 265 (Can.).

  103. . See Fla. Stat. § 794.011 (2017). The American Law Institute is working on a fourth model, one that includes a mental state for the offender that is included in the text of the statute. See, e.g., Model Penal Code § 213.6(1)(c) (Am. L. Inst., Tentative Draft No. 12, 2021).

  104. . See Estrich, supra note 3, at 1154–55.

  105. . Id. at 1154.

  106. . See infra Section III.A.

  107. . See Mich. Comp. Laws § 750.520b(1) (2017).

  108. . See id. § 750.520b(1)(c).

  109. . See id. § 750.520b(1)(d).

  110. . See id. § 750.520b(1)(e).

  111. . See Dressler, supra note 71, at 66.

  112. . Estrich, supra note 3, at 1154.

  113. . See Tuerkheimer, supra note 51, at 36.

  114. . See Estrich, supra note 3, at 1098–99.

  115. . See Criminal Code, R.S.C. 1985, c. C-46, s. 265 (Can.).

  116. . See John M. Burkoff & Russell L. Weaver, Inside Criminal Law: What Matters and Why 151–53 (2008).

  117. . Id. at 152.

  118. . See Estrich, supra note 3, at 1098–99.

  119. . See Schulhofer, supra note 34, at 102, 104.

  120. . See id. at 32.

  121. . See id. at 32, 96–97.

  122. . See id.

  123. . Fla. Stat. § 794.011(5)(b) (2017).

  124. . See id. § 794.011(1)(h), (5)(b).

  125. . See id. § 794.011(2)–(5).

  126. . See Dressler, supra note 71, at 66.

  127. . See id. at 154.

  128. . See Schulhofer, supra note 34, at 272–73. See generally Estrich, supra note 3, at 1133.

  129. . See Joint Serv. Comm. on Mil. Just., Manual for Courts-Martial United States pt. IV, art. 120(a), (b)(1) (2016), 10 U.S.C. § 920.

  130. . Id. at 120(b)(2)(A).

  131. . National Defense Authorization Act, H.R. 1815, 109th Cong. § 552(a)(1) (2006); see, e.g., Major Mark D. Sameit, When A Convicted Rape Is Not Really a Rape: The Past, Present, and Future Ability of Article 120 Convictions to Withstand Legal and Factual Sufficiency Review, 216 Mil. L. Rev. 77, 87–88 (2013).

  132. . See Criminal Code, R.S.C. 1985, c. C-46, ss. 265(1)(a)–(c), 269–73 (Can.).

  133. . Id. ss. 265(1)(a), (3).

  134. . Id. s. 273.1(2).

  135. . Bernard Schissel, Law Reform and Social Change: A Time-Series Analysis of Sexual Assault in Canada, 24 J. Crim. Just. 123, 123 (1996).

  136. . Criminal Code R.S.C. 1985, c. C-46, s. 273.2(a) (Can.).

  137. . Id. s. 273.2(a).

  138. . Id. ss. 273.2(b)–(c).

  139. . See Susan J. Lea et al., Attrition in Rape Cases: Developing a Profile and Identifying Relevant Factors, 43 Brit. J. Criminology 583, 583–84 (2003); see also Jeanne Gregory & Sue Lees, Attrition in Rape and Sexual Assault Cases, 36 Brit. J. Criminology 1, 14 (1996) (illustrating in table two that those moments are analyzed when studying the life cycle of a case).

  140. . See infra Section III.A.

  141. . See infra Section III.B

  142. . Julie Horney & Cassia Spohn, Rape Law Reform and Instrumental Change in Six Urban Jurisdictions, 25 L. & Soc’y Rev. 117, 128 (1991); see David John Frank et al., The Global Dimensions of Rape-Law Reform: A Cross-National Study of Policy Outcomes, 74 Am. Socio. Rev. 272, 274, 277 (2009).

  143. . Horney & Spohn, supra note 141, at 128.

  144. . Frank et al., supra note 141, at 277.

  145. . John K. Cochran et al., Deterrence or Brutalization? An Impact Assessment of Oklahoma’s Return to Capital Punishment, 32 Criminology 107, 118 (1994).

  146. . Jeff A. Bouffard & LaQuana N. Askew, Time-Series Analysis of the Impact of Sex Offenders Registration and Notification Law Implementation and Subsequent Modifications on Rate of Sexual Offenses, 64 Crime & Delinq. 1483, 1499 (2019).

  147. . James Lopez Bernal et al., The Use of Controls in Interrupted Time Series Studies of Public Health Interventions, 47 Int’l J. Epidemiology 2082, 2083–85 (2018).

  148. . Cassia C. Spohn & Julie Horney, The Impact of Rape Law Reform on the Processing of Simple and Aggravated Rape Cases, 86 J. Crim. L. & Criminology 861, 867 (1996).

  149. . Id. at 882, 884.

  150. . Rita Gunn & Rick Linden, The Impact of Law Reform on the Processing of Sexual Assault Cases, 34 Can. Rev. Socio. & Anthropology 155, 158–61 (1997).

  151. . James L. LeBeau, Statute Revision and the Reporting of Rape, 72 Socio. & Soc. Rsch. 201, 202 (1988).

  152. . Mimi Ajzenstadt & Odeda Steinberg, Never Mind the Law: Legal Discourse and Rape Reform in Israel, 16 Affilia 337, 344 (2001).

  153. . Wallace D. Loh, Q: What has Reform of Rape Legislation Wrought? A: Truth in Criminal Labelling, 37 J. Soc. Issues 28, 36, 40 (1981).

  154. . Clay-Warner & Burt, supra note 63, at 162–63.

  155. . Id. at 162.

  156. . Id. at 165.

  157. . Id.

  158. . Jeanne C. Marsh et al., Rape and the Limits of Law Reform 27 (1982).

  159. . Id. at 29.

  160. . Id. at 28–29, 38.

  161. . Id. at 38.

  162. . Id.

  163. . Schissel, supra note 134, at 123.

  164. . Id. at 126–27.

  165. . Julian V. Roberts & Robert J. Gebotys, Reforming Rape Laws: Effects of Legislative Change in Canada, 16 L. & Hum. Behav. 555, 561 (1992).

  166. . Id. at 561–63.

  167. . Id. at 568–570; see also Scott Clark & Dorothy Hepworth, Effects of Reform Legislation on the Processing of Sexual Assault Cases, in Confronting Sexual Assault: A Decade of Legal & Social Change 113, 125–26 (Julian V. Roberts & Renate M. Mohr eds., 1994) (finding that the reported change did not have any impact on the founding of sexual assault crimes or the decision to bring them to trial, although the founding rates for sexual assault crimes were nearly the same as that for all crimes of violence post-reform).

  168. . Kenneth Polk, Rape Reform and Criminal Justice Processing, 31 Crime & Delinq. 191, 194 (1985).

  169. . See id. at 193–94.

  170. . Id. at 195.

  171. . Id. at 196.

  172. . Id.

  173. . Id. at 197.

  174. . Horney & Spohn, supra note 141, at 117.

  175. . Id. at 126–27.

  176. . Id. at 129.

  177. . Id. at 149.

  178. . Ronet Bachman & Raymond Paternoster, A Contemporary Look at the Effects of Rape Law Reform: How Far Have We Really Come?, 84 J. Crim. L. & Criminology 554, 565 (1993).

  179. . Id. at 566.

  180. . Using data from Uniform Crime Reports (UCR) can be problematic. Until 2012, the UCR used the common law definition of rape (vaginal intercourse by a man to a woman through force and without consent). Because of this, this data has only included a small category of sexual assaults, which may be the ones that law enforcement will take most seriously, anyway.

  181. . Id.

  182. . Id. at 568.

  183. . Id. at 564–65.

  184. . Frank et al., supra note 141, at 278.

  185. . Id.

  186. . Id. at 279.

  187. . See id. at 284.

  188. . Eric R. Carpenter et al., The Effect of Changing the Military’s Sexual Assault Laws on Law Enforcement Investigative Findings in the US Army, 46 L. & Hum. Behav. 313, 313 (2022).

  189. . See id. at 316.

  190. . Id. at 317.

  191. . Id. at 319.

  192. . Ilene Seidman & Susan Vickers, The Second Wave: An Agenda for the Next Thirty Years of Rape Law Reform, 38 Suffolk U. L. Rev. 467, 468 (2005); Lisa Frohmann & Elizabeth Mertz, Legal Reform and Social Construction: Violence, Gender, and the Law, 19 L. & Soc. Inquiry 829, 850–51 (1995). See generally Caringella, supra note 64, at 57–58 (explaining that bias associated with rape reporting is still prevalent among legislative and executive actor).

  193. . Ajzenstadt & Steinberg, supra note 151, at 338.

  194. . Supra Section II.B.

  195. . Supra Section II.C.

  196. . Infra Part IV.

  197. . Supra Part II.

  198. . Supra Section II.C.

  199. . Supra Section II.C.

  200. . Supra Section II.C.

  201. . Supra Section II.C.

  202. . Criminal Jury Instructions & Model Colloquies: Credibility, NYCourts.Gov, [].

  203. . Albert Moore et al., Trial Advocacy: Inferences, Arguments, and Trial Techniques 70 (1996). See generally Dennis J. Devine, Jury Decision Making: The State of the Science 26–29 (2012).

  204. . See generally Giannelli, supra note 68, at 263–301.

  205. . In addition to attacking the credibility of the victim, the defendant can also elicit other facts that are helpful for the mistake of fact defense. The admissibility of those facts would be governed by the Due Process Clause, as described above.

  206. . Delaware v. Van Arsdall, 475 U.S. 673, 679 (1986) (emphasis added).

  207. . Id. at 680 (emphasis added).

  208. . The defense can also admit extrinsic evidence (coming from some source other than the cross-examination) that the victim consented. The admissibility of that evidence would be governed by rape shield rules and ultimately the Due Process Clause consistent with the discussion related to the mistake of fact defense.

  209. . Deborah Tuerkheimer, Judging Sex, 97 Cornell L. Rev. 1461, 1469–72 (2012).

  210. . Fed. R. Evid. 412(b)(1)(B).

  211. . Supra Section II.B, II.C.

  212. . Supra Section II.B, II.C.

  213. . See Dressler, supra note 71, at 155.

  214. . Chambers v. Mississippi, 410 U.S. 284, 294 (1973) (emphasis added).

  215. . Crane v. Kentucky, 476 U.S. 683, 690 (1986) (quoting California v. Trombetta, 467 U.S. 479, 485 (1984)) (emphasis added).

  216. . United States v. Burr, 25 F. Cas. 30, 37 (1807) (No. 14692D) (emphasis added).

  217. . Washington v. Texas, 388 U.S. 14, 19 (1967) (emphasis added).

  218. . This evidence could go beyond evidence that might be gained during the cross examination of the victim—the defendant can testify to it, as can other witnesses.

  219. . Dressler, supra note 71, at 144, 152.

  220. . Id. at 135.

  221. . Id. at 154.

  222. . Elonis v. United States, 575 U.S. 723, 734, 736 (2015) (quoting Carter v. United States, 530 U.S. 255, 269 (2000)).

  223. . See Dressler, supra note 71, at 580.

  224. . Id. at 143–45.

  225. . See generally Horney & Spohn, supra note 141, at 150.

  226. . See Schulhofer, supra note 34, at 254–73.

  227. . Criminal Code, R.S.C. 1985, c. C-46, s. 273.1(1) (Can.).

  228. . See Consent (Tate & Partners/Watch Out for the Bears Productions 2004).

  229. . See Supra Section II.A.

  230. . Joint Serv. Comm. on Mil. Just., Manual for Courts-Martial United States pt. IV, art. 120(g)(7)(A) (2016), 10 U.S.C. § 920.

  231. . See Caringella, supra note 64, at 208–09.

  232. . Criminal Code, R.S.C. 1985, c. C-46, s. 273.1(1.1) (Can.).

  233. . David P. Bryden, Redefining Rape, 3 Buff. Crim. L. Rev. 317, 325 (2000).

  234. . Joint Serv. Comm. on Mil. Just., Manual for Courts-Martial United States pt. IV, art. 120(g)(7) (2016), 10 U.S.C. § 920.

  235. . Supra Section II.C.

  236. . Model Penal Code § 213.0(2)(e)(ii) (Am. L. Inst., Council Draft no. 12, Dec. 2021). The American Law Institute (ALI) subsequently addressed the controversy surrounding its use of the word “inaction” in its April 2022 Tentative Draft of the Model Penal Code, but the word “inaction” still remains in the ALI’s most-recently adopted definition of consent, which may have a harmful effect on victims’ cases. See Model Penal Code, Rep.’s Memorandum xiv–xv (Am. L. Inst., Tentative Draft no. 6, Apr. 2022); see also Model Penal Code § 213.0(2)(e) (Am. L. Inst., Official Statutory Text, Sept. 2022).

  237. . Model Penal Code § 213.0(2)(e) (Am. L. Inst., Official Statutory Text, Sept. 2022).

  238. . Criminal Code, R.S.C. 1985, c. C-46, s. 273.1(2) (Can.).

  239. . Joint Serv. Comm. on Mil. Just., Manual for Courts-Martial United States, pt. IV, art. 120(g)(7) (2016), 10 U.S.C. § 920.

  240. . The ALI included similar language, but it required that the “no” be clear: “A clear verbal refusal—such as ‘No,’ ‘Stop,’ or ‘Don’t’—establishes the lack of consent.” Model Penal Code § 213.0(2)(e)(iv) (Am. L. Inst., Council Draft no. 12, Dec. 2021). This ignores what we know about how people communicate “no” and plays into the “mythcommunication.”

  241. . Criminal Code, R.S.C. 1985, c. C-46, s. 273.2 (Can.). This reform was passed in 1992, after the periods studied in the research discussed in Part III. See supra Part III.

  242. . Criminal Code, R.S.C. 1985, c. C-46, s. 273.2 (Can.).

  243. . See supra Section IV.A.

  244. . See Fed. R. Evid. 401.

  245. . See supra Part IV.

  246. . Fed. R. Evid. 412(c).

  247. . Id. at 412(b)(1)(B).

  248. . Giannelli, supra note 68, at 158.

  249. . Fed. R. Evid. 412(b)(1)(C).

  250. . Id. at 403.

  251. . See supra Section II.A.

  252. . See Tuerkheimer, supra note 51, at 57.

  253. . See Kristen M. Cuevas et al., Neurobiology of Sexual Assault and Osteopathic Considerations for Trauma-Informed Care and Practice, 118 J. Am. Osteopathic Ass’n e3 (2018); Cynthia V. Ward, Trauma and Memory in the Prosecution of Sexual Assault, 45 Law & Psychol. Rev. 87, 151–54 (2021).

  254. . Valliere, supra note 20, at 5.

  255. . See Fanflik, supra note 28, at 2; see also Jennifer Gentil Long, Explaining Counterintuitive Victim Behavior in Domestic Violence and Sexual Assault Cases, 1 Voice 1, 3 (2007).

Eric R. Carpenter

Please find a PDF copy of the article here.