Reorienting the Rules of Evidence

Latent heterosexism is deeply engrained in the common sensibility of the American legal system. In sexual assault cases, that latent heterosexism reveals itself in disparate admissibility rulings and practices regarding evidence that purports to demonstrate a defendant’s sexual orientation. When defendant and victim are of the same gender, evidence that a defendant is straight is used to prove innocence, while evidence that a defendant is gay or bisexual is used to prove guilt. This evidence is used to advance narratives invoking anti-gay and anti-bisexual stereotypes of deviance and criminality. Similar arguments are not made in cases involving opposite-gender defendant and victim. This disparate over-admission and misuse of evidence of defendant sexual orientation in same-gender sexual assault cases arise from a fundamental misunderstanding of both the nature of sexual orientation and the etiology of sexual assault. Although existing evidence doctrines provide a robust framework through which to properly exclude most sexual orientation evidence, chronic under-utilization of that framework calls for a reorientation of the rules of evidence towards a deeper understanding of sexual identity and behavior, and a more honest accounting of sexual prejudice and negative sexual stereotypes. This Article explores theoretical and scientific models of sexual orientation and sexual assault, demonstrating the significant attenuation between sexual orientation, desire, and assault. When properly grounded in those complex realities, the doctrines of logical relevance, character evidence, and the legal relevance balancing test operate to exclude most evidence of defendant sexual orientation in sexual assault cases. Crucially, this analysis must be highly cognizant of the powerful heterosexist forces that both overtly and implicitly bias fact-finders and militate against admission of this evidence in nearly all cases.

Introduction

Despite important progress towards equality for sexual minorities, pernicious anti-gay and anti-bisexual1This Article will use anti-gay to refer to that which is hostile towards same-gender sexual activity or intimate relationships, as well as gay and lesbian people, and anti-bisexual to refer to the unique hostility and discrimination faced by bisexual people, both because they are not heterosexual and because they do not conform to a binary categorization of sexual orientation. See, e.g., Mickey Eliason, Bi-negativity: The Stigma Facing Bisexual Men, 1 J. Bisexuality 137 (2000); Michele J. Eliason, The Prevalence and Nature of Biphobia in Heterosexual Undergraduate Students, 26 Archives Sexual Behav. 317 (1997); Gregory M. Herek, Heterosexuals’ Attitudes Towards Bisexual Men and Women in the United States, 39 J. Sex Res. 264 (2002); Tania Israel & Jonathan J. Mohr, Attitudes Toward Bisexual Women and Men, 4 J. Bisexuality 117 (2004); Patrick S. Mulick & Lester W. Wright Jr., The Biphobia Scale a Decade Later: Reflections and Additions, 11 J. Bisexuality 453 (2011). For further discussion of negative sexual stereotypes, see infra Section III.A. stereotypes,2Social scientists define stereotypes as “a set of beliefs about the personal attributes of a group of people.” See Angela Simon, The Relationship Between Stereotypes of and Attitudes Toward Lesbians and Gays, in Stigma and Sexual Orientation 62, 64 (Gregory M. Herek ed., 1998). sexual prejudice,3See Gregory M. Herek, Beyond “Homophobia”: Thinking About Sexual Prejudice and Stigma in the Twenty-First Century, 1 Sexuality Res. & Soc. Pol’y 6, 16 (2004) (citing Gregory M. Herek, The Psychology of Sexual Prejudice, 9 Current Directions Psychol. Sci. 19, 19 (2000)). Broadly conceived, sexual prejudice refers to negative attitudes based on sexual orientation whether their target is homosexual, bisexual, or heterosexual . . . . Given the power relations in contemporary society, however, prejudice is most commonly directed at people who engage in homosexual behavior or label themselves gay, lesbian, or bisexual. Id. and heterosexism4Id. at 15; see also Simon, supra note 2. persist in American society.5See generally Herek, supra note 3 (describing how these phenomena are sometimes referred to by the misnomer “homophobia”). A 2012 poll showed that forty-three percent of Americans believed “homosexual sex” was “always wrong.”6See Tom W. Smith & Jaesok Son, General Social Survey 2012 Final Report: Trends in Public Attitudes about Sexual Morality 1 (Apr. 2013), http://‌www.norc.org/‌PDFs/‌GSS%‌20Reports/‌Trends%‌20in%‌20Sexual%‌20Moraility_‌Final.pdf. Similarly, in 2015, a sizeable minority of Americans reported that they disapproved of same-sex sexual intimacy.7See Support for Same-Sex Marriage at Record High, but Key Segments Remain Opposed, Pew Res. Ctr. (June 8, 2015), http://‌www.people-press.org/‌2015/‌06/‌08/‌support-for-same-sex-marriage-at-record-high-but-key-segments-remain-opposed (indicating that thirty percent of Americans think homosexuality should be discouraged). Although acceptance of sexual minorities seemed to be growing in recent years, that trend appears to have reversed itself in 2017, when more non-LGBTQ Americans reported feeling uncomfortable around LGBTQ people than in prior iterations of the same annual poll.8          See Harris Poll, Accelerating Acceptance: A Survey of American Acceptance and Attitudes Toward LGBTQ Americans 1 (2018), https://‌www.glaad.org/‌publications/‌accelerating-acceptance-2018. Moreover, psychologists believe that sexual prejudice and negative sexual stereotypes can exist, on a subconscious level, even among those who purport to favor sexual equality.9See Ritch C. Savin-Williams et al., Sexual and Gender Prejudice, in 2 Handbook of Gender Research in Psychology 359 (Joan C. Chrisler & Donald R. McCreary eds., 2010) [hereinafter Sexual and Gender Prejudice]. Thus, even those who espouse values of equality and justice may be affected by subconscious heterosexist biases.

Negative implications of these heterosexist forces are manifested throughout our justice system.10See, e.g., Jude Hall, Lesbian Mothers & Gay Fathers: Overt and Subconscious Homophobic Biases Concerning Parenting Fitness, 1 Dalhousie J. Legal Stud. 53 (2010) (describing such biases within the context of child custody cases); Jennifer M. Hill, The Effects of Sexual Orientation in the Courtroom: A Double Standard, 39 J. Homosexuality 93 (2000) (discussing negative effects of sexual orientation in various justice system contexts). This Article will focus on one such manifestation: improper admission and overreliance upon evidence of defendant sexual orientation in sexual assault cases. My fundamental critique of judicial treatment of defendant sexual orientation in sexual assault cases is that it overtly benefits straight-perceived defendants and disadvantages defendants perceived as gay or bisexual. The most common arguments about defendant sexual orientation are: “Defendant is gay or bisexual, therefore he committed the sexual assault (on a man)” and “Defendant is straight, therefore he did not commit this sexual assault (on a man).”11See, e.g., Meny v. State, 861 S.W.2d 303, 306 (Ark. 1993); J.O. v. O.E., 100 A.3d 478, 480 (D.C. 2014); Moore v. State, No.      101,710, 2010 WL 3245293 (Kan. Ct. App. Aug. 13, 2010); Florence v. State, 755 So. 2d 1065, 1071 (Miss. 2000); State v. Caldwell, 662 S.E.2d 474, 482–83 (S.C. Ct. App. 2008); cf. People v. Hackett, 365 N.W.2d 120, 126 (Mich. 1984) (discussing evidence proffered by defendant showing complainant’s alleged homosexuality to establish defense of consent for rape charge). The inverse arguments—“Defendant is straight, therefore he committed sexual assault (on a woman)” and “Defendant is gay, therefore he did not commit sexual assault (on a woman)”—have not been made in reported cases.12Based on comprehensive review of reported cases using WestlawNext. See cases cited infra note 17. This disparity reflects underlying stereotypes and prejudices about gay, lesbian, and bisexual people, which can range from a basic notion that being gay or bisexual is abnormal and therefore evidentiarily noteworthy, to more pernicious conceptions of gay, lesbian, and bisexual people as hypersexual, deviant, or immoral.

Evidence of a defendant’s sexual orientation tells a fact-finder essentially nothing about whether he or she committed a sexual assault.13Nomenclature, definitions, and delineation of degrees of sex offenses vary widely by state. This Article will primarily use the term sexual assault to refer to any sexual contact without consent and considered unlawful in the pertaining jurisdiction and which state laws may refer to as, inter alia, rape (various), aggravated rape (Louisiana, Tennessee), simple rape (Louisiana), forcible rape (Louisiana), sexual assault (Colorado, Hawaii), sexual abuse (various), sexual battery (various), unlawful sexual conduct (Colorado, Maine), sodomy (various), and forcible sodomy (Utah). See State Rape Statutes, Am. Prosecutors Res. Inst., http://‌www.arte-sana.com/‌articles/‌rape_‌statutes.pdf (last visited Feb. 9, 2018); see also Laws in Your State, RAINN, https://‌rainn.org/‌statelaws (last visited May 26, 2018) (providing comprehensive information regarding each state’s sexual assault laws). Myriad complexities attenuate any possible link between this evidence and a conclusion about guilt. Worse, evidence of sexual orientation triggers a set of stereotypes and prejudices rooted in the dominant heteronormative American culture that results in a distortion of justice, privileging defendants who are perceived as heterosexual and gender-conforming, and disadvantaging both defendants and victims who fall outside of hetero- and cis-normative social constraints.14These negative effects can intersect with other forms of systemic and interpersonal oppression faced by particular groups, including, among others, people of color, transgender people, immigrants, individuals living in poverty, and people living with disability. In spite of these serious concerns, courts across the country routinely admit and rely upon sexual orientation evidence in deciding same-gender15Unless quoting or closely paraphrasing a source, this Article will use the term gender to mean an individual’s performance of predominantly “male” or “female” identity and/or a perception by others that an individual is “male” or “female.” This linguistic and conceptual choice represents a deviation from jurisprudence in this area, which primarily uses the term sex, either conflating biological sex and gender or ignoring the concept of gender altogether. The oversimplified notion of sex-as-gender reflects a dominant cultural misunderstanding of sex and gender as fully overlapping concepts, and of both sex and gender as binary traits. Differences along the sex and gender spectra further complicate the analysis presented in this Article and provide additional support for the idea that sexual orientation is more nuanced and less probative than it may seem. For discussion of the distinction between sex and gender, see Mary Anne C. Case, Disaggregating Gender from Sex and Sexual Orientation: The Effeminate Man in the Law and Feminist Jurisprudence, 105 Yale L.J. 1, 2 (1995) (“In women’s studies and related disciplines, however, the two terms have long had distinct meanings, with gender being to sex what masculine and feminine are to male and female.”). sexual assault cases, based on a fundamentally flawed understanding of both sexual orientation and sexual assault. Evidence that the defendant is gay or bisexual is presented as probative of guilt; evidence that he is heterosexual, probative of innocence.16Meny, 861 S.W.2d 303; J.O., 100 A.3d 478; Moore, 2010 WL 3245293; People v. Bouwman, No. 307325, 2014 WL 2351300 (Mich. Ct. App. May 29, 2014); Florence, 755 So. 2d 1065; Crawford v. State, 811 P.2d 67 (Nev. 1991); Kimberly v. State, 757 P.2d 1326, 1328 (Nev. 1988); State v. Swartsell, No. CA2002–06–151, 2003 WL 21998619 (Ohio Ct. App. Aug. 25, 2003); Caldwell, 662 S.E.2d 474. Evidence of a defendant’s sexual orientation often is presented in a context of coded language and innuendo, trading on stereotypes and prejudices against gay, lesbian, and bisexual people.17See, e.g., Bouwman, 2014 WL 2351300, at *1 (in which prosecutor argued “[i]t is reasonable to infer that defendant, a successful business man with a family, would not want anyone to know that he has sexually assaulted another man and would therefore invent a story about the incident”); Florence, 755 So. 2d at 1071–72 (holding that gay pornography found in defendant’s possession was relevant to defendant’s “capacity to commit a male-on-male sexual assault, something the vast majority of the population finds hard to conceive or envision”); Swartsell, 2003 WL 21998619, at *4 (in which prosecutor argued “the allegation of sex involved is male on male. That in itself, to a large extent, is not normal. Some reject it entirely” (internal quotation marks omitted)); State v. Jennings, No. 247, 1989 WL 102500, at *4 (Tenn. Crim. App. Sept. 7, 1989) (affirming conviction in which prosecutor argued, “So, what we do is appeal to your prejudice . . . . [I]s someone of this persuasion more likely to commit the kind of crime on a male victim, than a heterosexual or not. Now, if that’s prejudice, ladies and gentlemen, go ahead and they will be prejudiced”).

This Article examines why courts have found evidence of defendant sexual orientation in sexual assault cases particularly vexing and argues for an evidentiary approach that is firmly grounded in scientific and theoretical understandings of both sexual orientation and sexual assault. By properly applying existing evidentiary doctrines while staying conscious of heterosexist biases, courts can forge a fairer approach to admissibility questions, and reach more just and equitable outcomes. Character doctrine, in particular, offers a promising framework that has, thus far, been underutilized with respect to defendant sexual orientation in sexual assault cases. The unique benefit of a character analysis is that it strikes a balance between state power and defendants’ rights in a criminal setting through the mercy rule doctrine, which allows a defendant to weigh the risks of opening the door to his own character trait. Regardless of the legal framework used to analyze admissibility, courts must have a clear-eyed conceptualization of the explicit and implicit biases against gay and bisexual people that pervade American society, as well as a full appreciation of the staggering weight of heteronormativity in the justice system. This is perhaps the most profound challenge and the most promising source of transformation with respect to the evidentiary issues addressed in this Article.

My inquiry is focused on the specific issue of defendant sexual orientation in criminal and quasi-criminal cases,18The scope of this Article extends only to evidence of a defendant’s sexual orientation in criminal and quasi-criminal sexual assault cases, such as civil protection orders, which generally implicate criminal statutes.. but it is situated within a broader conversation about the evidentiary issues tied into sexual orientation in other justice system contexts. Evidence related to a victim’s sexual orientation is a worthy area of scholarly interest, but one that lies beyond the scope of this project. The promulgation of rape shield laws and subsequent vigorous debate as to their constitutionality is well-worn territory.19See generally Clifford S. Fishman, Consent, Credibility, and the Constitution: Evidence Relating to a Sex Offense Complainant’s Past Sexual Behavior, 44 Cath. U. L. Rev. 709 (1995) (arguing that rape shield laws are generally fair but additional legislative clarification is needed); Myka Held & Juliana McLaughlin, Rape & Sexual Assault, 15 Geo. J. Gender & L. 155, 171–178 (2014) (providing an overview of rape shield laws); Shawn J. Wallach, Rape Shield Laws: Protecting the Victim at the Expense of the Defendant’s Constitutional Rights, 13 N.Y.L. Sch. J. Hum. Rts. 485 (1997) (critiquing rape shield laws and arguing for the necessity of Supreme Court review); see also Cindy Ellen Hill, Chicken-Hawk!: Evidence of a Complainant’s Homosexuality Under Vermont’s Rape Shield Law, 22 Vt. L. Rev. 711 (1998) (arguing that rape shield laws should be narrowed to allow evidence of a complainant’s sexual orientation in a same-sex sexual assault prosecution). Rich scholarship has also emerged on the validity of defendant sexual orientation in Title VII litigation.20See generally Jessica A. Clarke, Inferring Desire, 63 Duke L.J. 525 (2013). However, there is a dearth of work on defendant sexual orientation in a criminal and quasi-criminal context. Without a strong and well-reasoned intellectual frame through which to examine this type of evidence, courts have taken ad hoc approaches producing inconsistent and often unfair results.

The goal of this Article is to advance the conversation in this area and provide a blueprint for courts to use in deciding whether evidence of defendant sexual orientation should be admitted in sexual assault cases. To that end, Part I provides an overview of the problem and demonstrates the scope of misuse of sexual orientation evidence in sexual assault cases through an explication of two exemplar cases: J.O. v. O.E. and State v. Ford.21100 A.3d 478 (D.C. 2014); 926 P.2d 245 (Mont. 1996). These cases illustrate how evidence of defendant’s sexual orientation can be used in discriminatory ways in different litigation contexts, and how the admissibility of that evidence is driven by stereotypes and conflated inferences. Part II examines the broad question of probative value of sexual orientation evidence as a threshold question on logical relevance as well as the foundation for further analysis under other principles of evidence law. This inquiry necessitates an exploration of theories and models of sexual orientation and sexual assault. Part III discusses risk of unfair prejudice inherent in sexual orientation evidence in this context, through the application of a legal relevance balancing test and character evidence rules, including the bar on propensity evidence. Counterintuitively, analogies to homicide cases involving a “gay panic” defense claim and child sex abuse cases provide some guidance towards conceptualizing a more rational application of evidentiary principles to sexual orientation evidence.22The “gay panic” defense or unwanted same-gender sexual advance defense refers to the claim, made by a homicide defendant, that the defendant killed the victim in self-defense against a same-gender sexual advance or assault by the victim. See Christina Pei-Lin Chen, Note, Provocation’s Privileged Desire: The Provocation Doctrine, “Homosexual Panic,” and the Non-Violent Unwanted Sexual Advance Defense, 10 Cornell J.L. & Pub. Pol’y 195, 201 (2000). Part IV proposes a vision for the application of existing evidentiary principles to ensure that sexual orientation evidence is excluded unless it has truly significant probative value that will outweigh both latent sexual prejudice and the risk that the fact-finder will overvalue the evidence based on a misunderstanding of sexual orientation and sexual assault.

I. Stereotypes in Action

A broad look at sexual assault cases across the country elucidates just how chaotic the jurisprudential landscape is on the issue of defendant sexual orientation.23See, e.g., Meny v. State, 861 S.W.2d 303, 306 (Ark. 1993); J.O., 100 A.3d at 480; Moore v. State, No. 101,710, 2010 WL 3245293, at *3–5 (Kan. Ct. App. 2010); Florence v. State, 755 So. 2d 1065, 1071 (Miss. 2000); State v. Caldwell, 662 S.E.2d 474, 482–83 (S.C. Ct. App. 2008). There is little consistency in approach from case to case. Of course, a court’s analysis will depend on a variety of factors—the precise nature of the evidence itself, which litigant offers the evidence, the proponent’s theory of relevance, whether the door has been opened by the other party, and other questions. However, even taking into consideration those variables, there is a lack of uniformity in the evidentiary principles applied and how those rules interact with the concept of sexual orientation. The only observable trend is that defendant sexual orientation is, over the last forty years, generally found to be admissible in sexual assault cases, particularly those involving adult victims, regardless of which evidentiary rule is applied or in what way.24See, e.g., Meny, 861 S.W.2d at 306; J.O., 100 A.3d at 480; Gunter v. State, 296 S.E.2d 622 (Ga. Ct. App. 1982); Moore, 2010 WL 3245293; Florence, 755 So. 2d at 1071; Ford, 926 P.2d 245; Caldwell, 662 S.E.2d at 482–83.

A closer look at specific cases reveals that courts are using problematic legal reasoning in resolving these evidentiary questions. Evidence of sexual orientation is being admitted in criminal sexual assault cases based upon a fundamental misunderstanding of sexual orientation and sexual assault, which will be explored more in Part II. First, to provide a better sense of the troubling reliance on stereotypes and the overall anti-gay and anti-bisexual tenor of these cases, I will discuss two examples: State v. Ford25926 P.2d 245. and J.O. v. O.E.26100 A.3d 478.

A. “Not All Members of Society Would Fit the Perpetrator’s Profile”

The case of State v. Ford provides a telling window into the mind of both a trial and appellate court, demonstrating strained legal reasoning and rampant stereotypes. This was a criminal prosecution for sexual assault in the state of Montana. Earl Dallas Ford, a bisexual man, was charged with sexual assault on an adult male victim, Brad Stahl.27Ford, 926 P.2d 245. The prosecution sought to prove that Ford committed a drug-assisted sexual assault on Stahl, who slept on the couch in Ford’s home after a barbecue.28Id. at 246–47. The State’s evidence consisted of the victim’s testimony, the victim’s cousin’s testimony (that the victim was not drunk on the night of the incident), and a bottle of baby oil (which purportedly corroborated part of the victim’s account, that the defendant had rubbed baby oil on him after the alleged assault).29Id.

The prosecution, on cross-examination, elicited testimony from the defendant that he was bisexual.30Id. at 248. Although defense counsel strenuously objected, the judge allowed the questioning to continue:

Q: It is true, is it not, that you have an interest in homosexuality?

A: Yes, it is.

[Defense counsel]: Your honor, I am going to object, that is so incredibly prejudicial. . . . .

THE COURT: Overruled.

Q: Would you explain that interest, please?

A: I just have an interest in men.

Q: Are you a homosexual, bisexual or what?

A: Bisexual.31Id. at 247–48.

Having succeeded in getting this testimony into evidence, the prosecutor then argued in closing, again over defense objection, “I am not gay bashing, I am not bisexual bashing, I have got better things to do. But we have got a man here . . . . that admitted he is bisexual. That’s exactly what he is charged with doing.”32Id. at 248.

But bisexuality was not “exactly what [Ford was] charged with doing.” In fact, defendant Earl Dallas Ford had been charged with committing sexual assault against a male victim, not with simply being bisexual. In suggesting that identifying as bisexual is equivalent to committing rape, the prosecutor’s conflation of sexuality and sexual assault rested on anti-bisexual stereotypes. This conflation highlights the fundamental problem with this kind of evidence: the risk that it will be misused in unfairly prejudicial ways.

At the conclusion of a jury trial, defendant was convicted and sentenced to ten years in prison.33Contrast Ford’s ten-year sentence to the six-month sentence given to Stanford University student athlete Brock Turner, who was convicted of sexually assaulting an incapacitated woman in 2016. The Turner case was factually similar to Ford in several ways: the victims were both incapacitated and had been at parties earlier on the evening of the incident. Key differences were (1) the gender of the victim and the presumptive sexual orientation of the assailant, and (2) that two eye witnesses saw Turner assaulting his victim. See Sam Levin, Ex-Stanford Swimmer Gets Six Months in Jail and Probation for Sexual Assault, Guardian (June 2, 2016), https://‌www.theguardian.com/‌us-news/‌2016/‌jun/‌02/‌stanford-swimmer-sexual-assault-brock-allen-turner-palo-alto; Michael E. Miller, All-American Swimmer Found Guilty of Sexually Assaulting Unconscious Woman on Stanford Campus, Wash. Post (Mar. 31, 2016), http://‌www.washingtonpost.com/‌news/‌morning-mix/‌wp/‌2016/‌03/‌31/‌all-american-swimmer-found-guilty-of-sexually-assaulting-unconscious-woman-on-stanford-campus. The Montana Supreme Court ignored this risk of unfair prejudice and affirmed the Ford ruling, holding that “the probative value of the State’s evidence [that defendant was bisexual] was not substantially outweighed by the danger of unfair prejudice.”34Ford, 926 P.2d at 250. The court clearly identified the potential for prejudice, noting, “[t]here will be, on virtually every jury, people who would find the lifestyle and sexual preferences of a homosexual or bisexual person offensive.”35Id. The court then shrugged off this risk, saying that Ford’s rights had not been violated by admission of the evidence.36Id. In an effort to explain why the potential for unfair prejudice was outweighed by the evidence’s probative value, the court suggested that only a certain type of person is capable of committing sexual assault on a member of the same gender: “Ford fit the profile of someone who would commit the act for which he was accused. This is especially true given the fact that, because of the nature of the crime, not all members of society would fit the perpetrator’s profile.”37Id.

This passage is rich with innuendo that betrays the underlying prejudices at work in the court’s reasoning: “the nature of the crime” was a same-gender sexual assault, and “the perpetrator’s profile” was, allegedly, a bisexual man.38This argument also implicates rules limiting the use of character evidence, as discussed in infra Section III.B. This language seems to suggest, with a wink, that we all know what that means. As the prosecutor had at trial, the appellate court conflated sexuality and sexual assault, failing to separate the logical inference that the defendant’s bisexuality could make him more likely to feel sexual attraction towards a man from the quite different inference that the defendant’s bisexuality would make him more likely to commit sexual assault. Instead, the court conflated these inferences, alluding to the notion that sex between two men, whether consensual or not, is aberrant—not something a supposedly normal person would do, as “not all members of society would fit the perpetrator’s profile.”39This sort of reference to sexual mores and sexual deviance with respect to a defendant’s sexual orientation in the context of sexual assault allegations is certainly not unique to this case. See State v. Swartsell, No. CA2002–06–151, 2003 WL 21998619, at *4 (Ohio Ct. App. Aug. 25, 2003) (in which prosecutor argued “the allegation of sex involved is male on male. That in itself, to a large extent, is not normal. Some reject it entirely” (internal quotation marks omitted)); see also Florence v. State, 755 So. 2d 1065, 1071–72 (Miss. 2000) (holding that evidence was relevant to defendant’s “capacity to commit a male-on-male sexual assault, something that the vast majority of the population finds hard to conceive or envision”).

B. “Since He Does Not Have a Homosexual Orientation, He Is Not Going to Approach [the Victim] for Sex”

The flawed logic surrounding sexual orientation and perpetration of sexual violence carries the risk of harm not only to defendants, but also to victims. The argument that because a defendant is heterosexual he could not have perpetrated sexual violence against another man conflates sexuality and sexual violence, just as the court did in Ford. The result is to deprive a victim of a fair access to justice by automatically indulging a problematic “I’m not gay” defense, even in a civil case. Such a theory rests on fundamentally flawed concepts of sexual orientation and sexual violence and affords undue privilege to a defendant who successfully (whether truthfully or not) presents himself as heterosexual.40Under my proposed approach, there is one key exception to a general bar on an “I’m not gay” defense in a same-gender sexual assault case, based on the mercy rule exception to the general bar on propensity character evidence. See infra Section IV.B.

In the case of J.O. v. O.E., for example, a District of Columbia trial court denied a civil protection order to a man who testified that he had been sexually assaulted multiple times by his male roommate.41J.O. v. O.E., 100 A.3d 478, 480 (D.C. 2014). Only two witnesses testified at trial: the victim and the alleged perpetrator.42Id. The victim, an immigrant man with a physical disability, testified the defendant had exposed himself and groped the victim, threatening him with immigration-related “trouble” if he refused these advances.43Id. The defendant testified that he was not gay and did not commit the assault.44Id.

The trial judge allowed, over objection, the defendant’s testimony that he was not gay.45Id. The court ultimately denied the protection order in a ruling that was based almost exclusively on that testimony.46Id. In the judge’s ruling he said that the defendant “steadfastly put [on] a strong defense that he is not gay.”47Id. (internal quotation marks omitted). The judge explained that the defendant’s sexual orientation was relevant because it meant he would not “approach the [victim] for sex.”48Id. (stating “since he doesn’t have a homosexual orientation, he is not going to approach [the victim] for sex”). But, the allegation was not that the defendant had approached the victim for sex; it was that the defendant had sexually assaulted the victim, groping him without consent while making threats. Here, the judge engaged in the same logical conflation as did the prosecutor and appellate court in State v. Ford: failing to decouple sexual orientation and sexual violence.

The D.C. Court of Appeals remanded the case, holding that the trial court had failed to provide a sufficient basis for its ruling and “may have” improperly relied upon defendant’s sexual orientation, which it said was “not substantially probative.”49Id. at 482. The appellate court described the trial judge’s logic as “simplistic and unsound,” noting that many sexual assaults against men are committed by men who identify as heterosexual.50Id. Although the court’s ruling was favorable to the petitioner, it applied a peculiar legal standard (“not substantially probative”) rather than existing evidentiary doctrine.51The standard “not substantially probative” does not have a firm root in American evidence law as a stand-alone concept. Perhaps the court meant that the probative value was substantially outweighed by the risk of unfair prejudice, pursuant to legal relevance, codified in the Federal Rules at Fed. R. Evid. 403. Nevertheless, in straying from established legal doctrine, the court may have created additional confusion in an already challenging area of law. The ruling also stopped short of adopting a hard line against the admissibility of evidence of sexual orientation,52This opinion left the door open for the judge to simply restate his problematic opinion in more palatable terms, such as a credibility determination.. thus, leaving open the possibility that such evidence will continue to be admitted in similar cases in the District of Columbia—either by judges inclined to more artfully disguise their prejudice or by juries, which are not required to provide an explanation for their findings.

II. Probative Values

Among the first questions at stake as to the admissibility of any piece of evidence is probative value. This is a concept understood relative to the proponent’s theory of relevance. Evidence must be probative of a material fact at issue through a permissible theory of relevance. Section B of this Part will examine the bedrock concept of logical relevance as applied to evidence of sexual orientation. Before reaching that inquiry, the question must first be asked: what is evidence of sexual orientation? Section A of this Part explores this question, revealing that the answer is not as simple as it may seem.

A. Evidencing Sexual Orientation

Assuming, for the time being, that sexual orientation is something definite and provable,53See infra Section II.B.1. for a discussion on the complex and ephemeral nature of sexual orientation. it is not necessarily immediately clear what constitutes evidence of sexual orientation. Looking again to the landscape of cases across the country, “sexual orientation evidence” can be broadly divided into two classes: items within the defendant’s possession54E.g., Florence v. State, 755 So. 2d 1065 (Miss. 2000); Felker v. State, 241 S.E.2d 576 (Ga. Ct. App. 1978); State v. Brown, No. 98-1987, 2000 WL 278548, at *5 (Iowa Ct. App. Mar. 15, 2000); State v. Creech, 495 S.E.2d 752, 755 (N.C. Ct. App. 1998). and various forms of testimony.55E.g., State v. Ford, 926 P.2d 245, 248 (Mont. 1996); see also J.O. v. O.E., 100 A.3d 478, 480 (D.C. 2014); Meny v. State, 861 S.W.2d 303, 306 (Ark. 1993). Proponents of this evidence, and the appellate judges who review it, are often vague about the evidence’s theory of relevance; so, one must look to how the evidence is used and what arguments it is marshaled to support in order to determine what the evidence is supposed to prove.

Tangible Items. Proof of possession of certain items, particularly gay pornography, has been a common form of evidence that arises in sexual assault cases where defendant sexual orientation is at issue. Evidence that a defendant possessed magazines, videos, and books has been offered in courts across the country.56See Florence, 755 So. 2d 1065; Felker, 241 S.E.2d 576; Brown, 2000 WL 278548. For example, in the case of Florence v. State, pornography featuring sex acts between two men that allegedly had been found in the defendant’s possession was admitted into evidence.57See Florence, 755 So. 2d 1065. The court found that the evidence was relevant to defendant’s “capacity to commit a male-on-male sexual assault, something the vast majority of the population finds hard to conceive or envision.”58Id. at 1071–72. While a male defendant’s possession of sexually explicit materials featuring two men engaged in sex acts may evidence sexual desire towards the opposite sex, it certainly does not evidence proclivity towards sexual violence.59See infra Section III.C.

In other cases, possession of a particular item is used not to prove a general tendency towards homosexuality or sexual violence, but some other material fact. For example, a defendant’s possession of a specific type of pornography has been offered to corroborate a victim’s testimony about an assault that involved showing the victim that particular type of pornography.60See Brown, 2000 WL 278548, at *5; State v. Creech, 495 S.E.2d 752, 755 (N.C. Ct. App. 1998). This theory of relevance is significantly more sound than one premised on blatant stereotypes; and the evidence may be highly probative in a corroborative function. However, it is often difficult if not impossible to separate the improper use of the evidence from this narrow proper inference. Whether a legal relevance (probative value versus risk of unfair prejudice) balancing test is a sufficient safeguard in such an instance is the subject of further discussion later in this Article.

Testimony. Perhaps the most common form of evidence regarding defendant sexual orientation in sexual assault cases is testimony. Within the broad class of testimony exist a few key distinctions. The first is between the defendant’s own testimony and third-party testimony. The second is between testimony about identity (self-identified or perceived by others) or behavior. For example, in Ford, the defendant himself testified that he was bisexual and that he had “an interest in homosexuality,”61See State v. Ford, 926 P.2d 245, 248 (Mont. 1996). and in J.O., the defendant testified that he was not gay.62See J.O. v. O.E., 100 A.3d 478, 480 (D.C. 2014). Testimony can also be about defendant’s past behavior, such as relationships with men or women,63See, e.g., Meny v. State, 861 S.W.2d 303, 306 (1993). watching a particular kind of pornography,64See Peter Nicolas, “They Say He’s Gay”: The Admissibility of Evidence of Sexual Orientation, 37 Ga. L. Rev. 793, 813–19 (2003). or visiting a particular type of establishment, such as a gay club.

There are several potential pitfalls with testimony about a defendant’s sexual orientation. Although a defendant would have personal knowledge of his or her own self-identity, that testimony is likely to be self-serving. A third party may be more objective but cannot have direct personal knowledge of a defendant’s self-identity. If a third party can testify about his or her perception of a defendant’s identity, it is not at all clear that such testimony is reliable or probative. Does the fact that someone else perceived the defendant to be gay matter? Perhaps not without additional meaningful testimony about why the witness holds that perception. That leads into a more promising form of third party testimony about a defendant’s past behavior, such as the fact that he was previously in a relationship with a man. This may be more trustworthy and concrete, but Section B will unsettle assumptions that such evidence is truly probative of a material fact at issue in a sexual assault case. Furthermore, as Part III argues, even if such testimony were probative, it represents propensity evidence, which is generally inadmissible under the Federal Rules of Evidence and the common law character evidence doctrine.

B. Logical Relevance

As a first step towards understanding if and when evidence of a defendant’s sexual orientation should be admissible in a sexual assault claim, this Section will explore the notion of logical relevance, which requires an inquiry into the probative value of such evidence. Closer examination of the construct of sexual orientation as well as our current understanding of sexual assault reveals that a defendant’s sexual orientation evidence has, at best, limited probative value in a sexual assault case; and, in some circumstances, it has no probative value at all. While it is not possible to rule out that this evidence could ever be relevant, its probative value is always profoundly limited.

All evidence must survive the basic standard of logical relevance in order to be admissible.65See Fed. R. Evid. 402; Paul R. Rice & Roy A. Katriel, Evidence: Common Law and Federal Rules of Evidence 67–68 (6th ed. 2009). The common law notion of logical relevance (simply called “relevance” under the federal rules) is succinctly defined in Federal Rule of Evidence (FRE) 401 as having “any tendency to make a fact [of consequence] more or less probable than it would be without the evidence.”66Fed. R. Evid. 401. Courts and scholars have emphasized that “any tendency” is a low standard requiring very little probative value.67McCormick on Evidence § 185 (Kenneth S. Broun ed., 7th ed. 2014); 29 Am. Jur. 2d Evidence § 306 (2008) (citing State v. Miller, 676 S.E.2d 546 (N.C. Ct. App. 2009), review denied, 683 S.E.2d 216 (N.C. 2009)) (“To be relevant, evidence need not bear directly on the question in issue if it is helpful to understand the conduct of the parties or their motives or if it reasonably allows the jury to draw an inference as to a disputed fact; the value of the evidence need only be slight.”). Logical relevance demands a logical relationship between a piece of evidence and the inference it is offered to support. As one court put it, a given piece of evidence is logically relevant if it “logically, naturally, and by reasonable inference” tends to make a material fact at issue in the litigation more or less likely to be true.68People v. Wilson, 136 P.3d 864, 869 (Cal. 2006) (citing People v. Harris, 118 P.3d 545 (Cal. 2005)) (internal quotation marks omitted).

1. Sexual Identity, Desire, and Behavior

Widespread misconceptions about sexual orientation confuse issues of sexual identity and sexual behavior, which leads fact-finders to misinterpret and overestimate the value of sexual orientation evidence. Sexual orientation is incredibly complex and poorly understood in dominant cultural and legal discourse. But a solid understanding of sexual orientation is required in order to properly analyze both whether a given piece of evidence proves that a defendant is of a particular sexual orientation, and whether a defendant’s sexual orientation proves any fact at issue in a sexual assault case. With a better understanding of sexual orientation, it becomes clear how attenuated the link between a piece of evidence purported to indicate sexual orientation and any meaningful conclusion really is.

Sexual orientation is a concept that describes a cluster of phenomena, including identity, behavior, romantic love, sexual attraction, and desire.69The extent to which sexual orientation is rooted in a biological basis remains an unresolved but increasingly irrelevant question. For years, an essential inquiry in the debate on gay rights was: Is sexual orientation immutable? Today, we recognize that LGBTQ rights should be protected regardless of the answer to that question. Although activists believed that proving the immutability of sexual orientation would inspire more Americans to support gay rights, it is unclear whether that belief was correct. See, e.g., M.K.B. Darmer & Tiffany Chang, Moving Beyond the “Immutability Debate” in the Fight for Equality After Proposition 8, 12 Scholar 1, 2–3 (2009); Janet E. Halley, Sexual Orientation and the Politics of Biology: A Critique of the Argument from Immutability, 46 Stan. L. Rev. 503, 506 (1994); Kathleen Hubbard & Peter Hegarty, Why Is the History of Heterosexuality Essential? Beliefs About the History of Sexuality and Their Relationship to Sexual Prejudice, 61 J. Homosexuality 471 (2014) (citing G.B. Lewis, Does Believing Homosexuality Is Innate Increase Support for Gay Rights?, 37 Pol’y Stud. J. 669 (2009)). Each of these phenomena can operate independently and each may change over time. In the field of human sexuality, sexual orientation is generally defined as “the sexual attraction, identity, arousals, fantasies, and behavior individuals have for one sex, the other sex, or both sexes.”70Zhana Vrangalova & Ritch C. Savin-Williams, Mostly Heterosexual and Mostly Gay/Lesbian: Evidence for New Sexual Orientation Identities, 41 Archives Sexual Behav. 85, 85 (2012). One model that neatly captures (and simplifies) the multiple dimensions of sexual orientation is psychologists Zhana Vrangalova and Ritch Savin-Williams’s theory of three components of sexual orientation: identity, romantic or sexual attraction, and behavior.71Ritch C. Savin-Williams, Who’s Gay? Does It Matter?, 15 Current Directions Psychol. Sci. 40, 41 (2006).

Contrary to commonly held belief, these phenomena—identity, attraction, and behavior—are not always closely aligned.72See Vrangalova & Savin-Williams, supra note 70, at 86 (outlining “[d]ecades of research” on the variances between individuals’ sexual identity, attraction, and behavior). The reasons for this are not well-understood, but it appears to be a function of normal expressions of human sexuality. Research has shown that a subset of individuals who identify as lesbian or gay report opposite-gender attraction or behavior, while some heterosexual-identified individuals report same-gender attraction or behavior.73          Id. at 86, 94 (citing Lisa Diamond, Sexual Fluidity: Understanding Women’s Love and Desire (2008)); Margaret Rosario et al., Sexual Identity Development Among Lesbian, Gay, and Bisexual Youths: Consistency and Change Over Time, 43 J. Sex Res. 46 (2006); Ritch Savin-Williams, The New Gay Teenager (2005). Some people may hide their sexual orientation (or remain “in the closet”) for many reasons, including an ongoing process of self-discovery, the evolving nature of their own sexual orientation, and a high value on privacy, as well as the possibility of significant social or psychological cost to openly identifying as gay, lesbian, or bisexual. Such individuals’ identities would not match their attraction and may or may not match their behavior at different points in time. Others may experiment with various sexual behaviors or simply not fit into an absolutely categorical model of sexual orientation. Research indicating that sexual orientation is fluid and can actually change over time adds another level of complexity to the concept of sexual orientation.74See, e.g., Lisa M. Diamond, Sexual Fluidity: Understanding Women’s Love and Desire (2008) (presenting findings of a longitudinal study on nearly 100 women that sexual orientation is not static). Differences in gender identity and performance, including transgender identity, further complicate commonly held assumptions about sex, gender, and sexual orientation.75See supra note 15.

Because of these normal variances in human sexuality, the probative value of sexual identity or past sexual behavior76Desire is not mentioned in this analysis because of the practical challenges associated with producing evidence of sexual desire. Instead, sexual desire is generally argued as an inference based upon behavior, e.g., possession of a certain type of pornography. See, e.g., Florence v. State, 755 So. 2d 1065 (Miss. 2000); State v. Young, No. 102,121, 2013 WL 6839328 (Kan. Ct. App. Dec. 27, 2013); State v. Tizard, 897 S.W.2d 732, 744–45 (Tenn. Crim. App. 1994). as to any specific instance of conduct (i.e., whether a defendant committed a sexual assault) is, at best, extremely limited. For example, in the J.O. v. O.E. case discussed above, even assuming that defendant’s testimony that he identified as heterosexual was truthful, he may have had same-gender sexual partners in the past and he may have felt some level of same-gender sexual desire, even while identifying as heterosexual. Similarly, even if the defendant in State v. Ford identified as bisexual and had, as he testified, “an interest in men,” he may have never had a male sexual partner before, nor gone on to have one in the future. Thus, in both cases the potential probative value of the testimony about the defendants’ sexual orientation is low.

Sexual orientation has limited evidentiary value not only because it is multi-faceted and dynamic, but also because it exists along a broad spectrum, beyond simple categories of heterosexual, bisexual, and gay/lesbian. The understanding of sexual orientation as a spectrum was made famous in the mid-twentieth century by researcher Alfred Kinsey and his collaborators, who developed the Kinsey Scale to describe their research findings.77Alfred C. Kinsey et al., Sexual Behavior in the Human Male 636–39 (1948) [hereinafter Kinsey, Sexual Behavior in the Human Male]; Alfred C. Kinsey et al., Sexual Behavior in the Human Female 468–74 (1953). The Kinsey Scale classifies sexual orientation on a spectrum from zero to six with zero representing “exclusively heterosexual” and six representing “exclusively homosexual.”78Kinsey, Sexual Behavior in the Human Male, supra note 77, at 638. Building upon Kinsey’s work, psychologists have subsequently posited three-dimensional models in an attempt to capture other facets of sexual orientation such as intensity of sexual attraction and change over time. See, e.g., Fritz Klein, The Bisexual Option (2d ed. 1993); Michael D. Storms, Theories of Sexual Orientation, 38 J. Personality & Soc. Psychol. 783, 784 (1980). Recent research supports the hypothesis that sexual orientation is actually continuously distributed along a spectrum, meaning that individuals fall at infinitely many different points along the spectrum, defying finite categorization.79Vrangalova & Savin-Williams, supra note 70, at 86.

These complexities fail to translate into a courtroom context because, while social scientists generally agree that sexual orientation exists along a continuum, mainstream discourse still favors a hardline system of classification that divides individuals into three discrete categories of heterosexual, bisexual, and gay/lesbian.80Even social science researchers have used this tripartite classification scheme for many years, which continues to inaccurately shape our understanding of human sexuality. Id. at 99. Like both J.O. v. O.E. and State v. Ford, sexual assault cases across the country reflect an unexamined dominance of a three-part categorization of sexual orientation. Litigants and witnesses are referred to as “heterosexual/straight,” “homosexual/gay/lesbian,” or “bisexual,” as though those classifications are the only options.81See, e.g., Meny v. State, 861 S.W.2d 303 (Ark. 1993); J.O. v. O.E., 100 A.3d 478 (D.C. 2014); Moore v. State, No. 101,710, 2010 WL 3245293 (Kan. Ct. App. Aug. 13, 2010); People v. Bouwman, No. 307325, 2014 WL 2351300 (Mich. Ct. App. May 29, 2014), appeal denied, 857 N.W.2d 43 (Mich. 2014); Florence, 755 So. 2d 1065; Crawford v. State, 811 P.2d 67 (Nev. 1991); Kimberly v. State, 757 P.2d 1326, 1328 (Nev. 1988); State v. Swartsell, No. CA2002–06–151, 2003 WL 21998619 (Ohio Ct. App. Aug. 25, 2003); State v. Caldwell, 662 S.E.2d 474 (S.C. Ct. App. 2008). There is no judicial recognition that these particular orientations may not mean the same thing to everyone, that these identities will not always match attraction and behavior, or that they may change over time.

A three-part classification of sexual orientation is an inherently flawed oversimplification; it misrepresents scientific understanding of human sexuality. When used in a judicial setting, this oversimplification enables faulty fact-finding, and undermines the interests of justice. For example, anyone who falls in the “mostly straight” or “mostly gay/lesbian” categories, which Vrangalova argues is a significant number of people, cannot be accurately captured by a three-part classification model.82Vrangalova & Savin-Williams, supra note 70, at 86. A defendant who testifies that he is straight (assuming he is being truthful) may be choosing the socially defined category that is closest to his actual orientation because he is what Vrangalova calls “mostly straight”—a Kinsey one, or two. For this hypothetical defendant, “straight” may be the best fit of the three options available in mainstream society and the courts. There may be significant pressure to fit into the category of heterosexual. Meanwhile, a fact-finder who hears that defendant is “straight,” like the judge in J.O. v. O.E., is likely to interpret that to mean he feels desire for and engages in sexual behavior exclusively with women. However, in many instances, this inference will be inaccurate. Thus, evidence of sexual orientation is particularly susceptible to misinterpretation by fact-finders.

Another important problem with sexual orientation evidence is the insufficiency of any given piece of evidence to truly prove an individual’s sexual orientation. There is a logical gap between what is offered as evidence of sexual orientation and a person’s actual complex multifaceted sexual orientation. A better approach would be to parse those facets and bring precision to the analysis by separating evidence of sexual identity, sexual desire, and sexual behavior. A statement, as in Ford, from the defendant that he is bisexual is evidence of sexual identity (not sexual desire, nor behavior). Indications of defendant’s past sexual relationships with men would be evidence of sexual behavior,83See Meny, 861 S.W.2d at 306. as would a particular type of pornography found in defendant’s possession.84See Florence, 755 So. 2d at 1071. These are ways to indicate the past performance of sexual behaviors, but again, evidence of behavior is not necessarily evidence of desire or identity. Therefore, to properly assess admissibility, courts must precisely identify and limit the scope of any evidence offered to prove an individual’s sexual orientation.

2. Etiologies and Models of Sexual Assault

In order to assess the probative value of evidence of a defendant’s sexual orientation in a sexual assault case, it is important to understand that both the cause (etiology) and the nature (model) of sexual assault are deceptively complicated. A deeper understanding of the etiology and nature of sexual assault is an essential precursor to proper analysis of sexual orientation evidence. The intricate and nuanced landscape of rape etiologies and models significantly attenuates the link between sexual orientation and sexual assault. At the same time, these complexities and nuances also make it difficult to conclude that a defendant’s sexual orientation could never be logically relevant in a sexual assault case. While it is clear that rape is a crime of power, it is not clear that sexual desire is never a factor in any rape.

Rape etiology is an evolving area of theorization among social scientists and scholars.85See generally Susan Faupel, Chapter 2: Etiology of Adult Sexual Offending, Off. Sex Offender Sent’g, Monitoring, Apprehending, Registering, & Tracking, https://‌www.smart.gov/‌SOMAPI/‌sec1/‌ch2_‌etiology.html (last visited May 26, 2018) (describing many different theories of sexual assault etiology); Beverly A. McPhail, Feminist Framework Plus: Knitting Feminist Theories of Rape Etiology into a Comprehensive Model, 17 Trauma, Violence & Abuse 314, 321 (2015) (critiquing and synthesizing, or “knitting,” various theories of rape etiology (citation omitted) (internal quotation marks omitted)); Jessica A. Turchik et al., An Examination of the Gender Inclusiveness of Current Theories of Sexual Violence in Adulthood: Recognizing Male Victims, Female Perpetrators, and Same-Sex Violence, 17 Trauma, Violence & Abuse 133 (2015) (examining gender inclusivity of various theories of rape). The most prominent theory, initially popularized by Susan Brownmiller,86McPhail, supra note 85. has been called the “patriarchal power and control” theory.87     Id. at 317 (citation omitted) (internal quotation marks omitted). According to this theory, rape is a function of power and hostility, rather than sex or desire.88Id. Originally conceptualized by radical feminists, the patriarchal power and control theory of rape gained traction in mainstream women’s movement in the 1970s and has remained dominant among mainstream feminists ever since.89Id. at 316 (noting that the patriarchal power and control theory of rape has remained largely free from critique). Under this theory, rape is thought to be motivated by power and control, as opposed to sexual desire.90Id. at 314–17; see Susan Brownmiller, Against Our Will: Men, Women and Rape (1975); Lilia Melani & Linda Fodaski, The Psychology of the Rapist and His Victim, in Rape: The First Sourcebook for Women 82, 84–85 (Noreen Connell & Cassandra Wilson eds., 1974). Similarly, rape is seen as an act of violence, not sex.91McPhail, supra note 85, at 314–18. The theory was supported by Nicholas Groth’s then groundbreaking 1979 research, which concluded:

[C]areful clinical study of offenders reveals that rape is in fact serving primarily nonsexual needs. It is the sexual expression of power and anger. Forcible sexual assault is motivated more by retaliatory and compensatory motives than by sexual ones. Rape is a pseudosexual act, complex and multidetermined, but addressing issues of hostility (anger) and control (power) more than passion (sexuality).92A. Nicholas Groth & H. Jean Birnbaum, Men Who Rape: The Psychology of the Offender 2 (1979).

The patriarchal power and control theory of rape has been transformational, and still holds significant value in understanding sexual assault today. It is etched in stark relief against a longstanding dominant culture of victim-blaming and pernicious rape myths, and represents a critical step towards more just treatment of rape victims. As one writer put it, “[t]he theoretical shift to view rape as motivated by power instead of sex played an important role in shifting blame away from female victims, and as a consequence, the physical attractiveness and sexual history of rape survivors became less relevant (although clearly vestiges of the practice remain today).”93McPhail, supra note 85, at 316. This shift was vital in developing a better understanding of sexual assault and helping victims find justice in a time when women were seen as responsible for their rapes, male lust was seen as unstoppable, and rape was seen as a natural consequence of men and women left alone together.94Notably, President Donald Trump seems to share the view that rape is a natural consequence of men and women alone together, tweeting in 2013, “26,000 unreported sexual assults [sic] in the military-only 238 convictions. What did these geniuses expect when they put men & women together?” Donald J. Trump (@realDonaldTrump), Twitter (May 7, 2013, 4:04 PM), https://‌twitter.com/‌realdonaldtrump/‌status/‌33190738‌3771‌148288?‌lang=‌en. When questioned about this tweet during the presidential campaign in 2016, Trump stated that it was “absolutely correct.” Tessa Berenson, Donald Trump Defends Tweet About Military Sexual Assault, Time (Sept. 8, 2016), http://‌time.com/‌4483354/‌donald-trump-military-sexual-assault-forum. The patriarchal power and control theory of rape also helped contextualize sexual violence as a gender-based crime, part of a broader patriarchal framework that served to uphold male power over women.95McPhail, supra note 85, at 316–17.

Despite its social importance and continued explanatory power, the patriarchal power and control theory has limitations. The theory has been criticized as oversimplified and too focused on cases of male-on-female rape by physical force, ignoring same-gender rape and various forms of sexual coercion.96Id. at 317. Even the social science originally used to support the patriarchal power and control theory of rape is nuanced—indicating that some rape seems to be motivated at least in part by sexual desire. While the gist of Groth’s much-cited research on this subject is that most rape is about power and anger, he includes several critical caveats: Groth says that “forcible sexual assault” (as opposed to what he characterizes as coercive sexual assault) “is motivated more by retaliatory and compensatory motives than by sexual ones,” but he does not rule out sexual motives.97Compare Groth, supra note 92, with A. Nicholas Groth & Ann Wolbert Burgess, Male Rape: Offenders and Victims, 137 Am. J. Psychiatry 806, 809 (1980) (explaining that some sexual offenders have unresolved conflicting sexual interests due to insecurity about their sexual orientation or sexual identity). While the patriarchal power and control theory of rape rightly calls into question assumptions about the link between sexual desire and sexual assault, it is does not fully account for the entire range of human behavior.

Other theories attempt to more fully capture the multiplicity of motivations and experiences of sexual assault.98See generally McPhail, supra note 85, for a comprehensive overview of models of sexual assault, including what she calls Catharine MacKinnon’s “normative heterosexuality perspective” (citing Catharine A. MacKinnon, Toward a Feminist Theory of the State (1989)) and intersectional models of rape envisioned by Angela Y. Davis and Kimberlé Crenshaw (citing Angela Y. Davis, Women, Race & Class (1981); Kimberlé Crenshaw, Mapping the Margins: Intersectionality, Identity Politics, and Violence Against Women of Color, 43 Stan. L. Rev. 1241 (1991)). Professor Elizabeth M. Iglesias proposes three distinct but potentially overlapping conceptual models for understanding of rape—“rape as hate crime,” “rape as sex,” and “rape as power”—and argues that excessive emphasis on “rape as hate crime” has limited our societal capacity to understand and respond to rape:

Rape as Hate Crime: . . . The rapist is a deviant person acting out his anti-social hostility on any unlucky woman unfortunate enough to happen across his path at the wrong time, in the wrong place. . . . . Enormous amounts of cultural resources are deployed to maintain this image as the dominant account of what rape is. Nevertheless, this account obscures the many other psycho-social contexts and manners in which women’s sexual autonomy is assaulted by men.99Elizabeth M. Iglesias, Rape, Race, and Representation: The Power of Discourse, Discourses of Power, and the Reconstruction of Heterosexuality, 49 Vand. L. Rev. 869, 891 (1996).

In this description of the “rape as hate crime” model, Iglesias points out the danger of hyper-focus on this one model of rape and encourages a wider interrogation of sexual power imbalances.

Iglesias’s “rape as hate crime” and “rape as power” models both largely track with the dominant patriarchal power and control theory of rape, representing the many instances of rape that are not primarily sexual in nature. However, Iglesias also offers a model of “rape as sex,” which adds to the discourse a framework for understanding some rape as both an act of sex and an act of violence:

Rape as Sex: In this account, rape reflects the fact that women’s consent is simply not relevant to some, many, or most men’s sexual gratification. . . . . Because [men] are ordinarily bigger and stronger than the women they are with, they take [sex] with more or less force and little awareness of the woman’s lack of consent. In these accounts, rape is not necessarily intentional, but the sex usually involves at least some degree of coercion and physical force.100Id. at 892.

This model tracks more closely with the common conception of acquaintance-rape (or “date-rape”) and sexual coercion, wherein rape occurs when power is used to serve the perpetrator’s desire. However, some believe rape, sex, and power are even more closely linked. Most prominently, Catharine MacKinnon theorized that rape exists on a continuum with heterosexual sex,101MacKinnon, supra note 98, at 171–72. suggesting that rape can be inextricably connected with eroticism and sexual desire on the part of the perpetrator, and that domination can be a central feature all along that continuum.

Under a strict interpretation of the patriarchal power and control theory of rape, a defendant’s sexual orientation as a proxy for desire or sexual behavior would never be relevant to whether that defendant committed a particular sexual assault. The evidence would always be excluded as logically irrelevant because it has no tendency to prove that the defendant committed the alleged crime of sexual assault. Although such an approach is appropriate in many sexual assault cases, it may miss some of the messy complexities of human behavior in some cases. The theories of rape discussed above call for a more nuanced approach that contemplates the complex and multi-dimensional aspects of sexual assault and may require judges to rely on the balancing test of legal relevance in sexual assault cases where desire cannot be ruled out as part of the crime itself.

Ultimately, we have an imperfect understanding of what motivates sexual assault. Empirical study of the motives underlying sexual assault is virtually impossible due to the significant challenges associated with examining something so tangled, private, and stigmatized. The existing evidence and current theory seem to indicate that although power is one of its central features, sexual assault is a complex heterogenous phenomenon that cannot be easily distilled to a single motive applicable in every case. Still, under any theory or model of rape discussed above, the probative value of a defendant’s sexual orientation is, at most, highly limited. Desire is only one piece of the puzzle in some cases; and, even where desire may be relevant, as discussed in Section II.B above, sexual orientation does not reliably predict desire or behavior.

III. Unfairly Prejudiced?

Logical relevance alone does not render a given piece of evidence admissible. Even when evidence is probative, the rules of evidence impose additional barriers. In essence, the pertinent questions are whether the evidence is fair and whether it will be more helpful than distracting to the fact-finder. These questions are organized into the doctrines of legal relevance and character evidence. What follows is a discussion of those evidentiary doctrines and the fairness and helpfulness questions that underlie them as applied to evidence of defendant sexual orientation in sexual assault cases. This Part ends with an examination of analogous case types, “gay panic” homicide and child sex abuse cases, which facilitate a reimagining of the application of existing rules of evidence to defendant sexual orientation evidence in adult sexual assault cases.

A. Legal Relevance

Common law and codified rules of evidence have come to recognize that certain evidence, while probative, should not be admissible because it carries too high a risk of misuse. This is the genesis of both the bar on propensity character evidence,102For detailed discussion of this concept, see infra Section III.B. and the general catch-all concept of legal relevance (balancing probative value and prejudicial effect), both of which can operate to exclude evidence of a defendant’s sexual orientation in sexual assault cases.

Otherwise admissible evidence is subject to discretionary exclusion under the catch-all legal relevance rule.103Fed. R. Evid. 403. The concept of legal relevance, sometimes called the prejudice rule,104See Andrew K. Dolan, Rule 403: The Prejudice Rule in Evidence, 49 S. Cal. L. Rev. 220 (1976). balances the benefit and detriment to the fact-finder of a given piece of evidence, granting judges the discretion to exclude problematic evidence on a case-by-case basis.105Paul R. Rice & Roy A. Katriel, Evidence: Common Law and Federal Rules of Evidence 75 (5th ed. 2005). Federal Rule of Evidence 403 represents the legal relevance approach that has been widely adopted by most states: “The court may exclude relevant evidence if its probative value is substantially outweighed by a danger of one or more of the following: unfair prejudice, confusing the issues, misleading the jury, undue delay, wasting time, or needlessly presenting cumulative evidence.”106Fed. R. Evid. 403.

Unfair prejudice requires more than simple damage to a party’s case; all evidence is offered to support the proponent’s case and damage the opposing party’s case.107E.g., Ayala v. State, 923 A.2d 952 (Md. Ct. Spec. App. 2007), cert. denied, 931 A.2d 1095 (Md. 2007); People v. Vasher, 537 N.W.2d 168, 171 (Mich. 1995); State v. Fisher, 828 N.W.2d 795, 800 (S.D. 2013); State v. Blevins, 744 S.E.2d 245, 260 n.10 (W. Va. 2013). As the FRE Advisory Committee explains, “unfair prejudice” means “undue tendency to suggest decision on an improper basis, commonly, though not necessarily, an emotional one.”108Fed. R. Evid. 403 advisory committee’s note. In same-gender sexual assault cases, unfair prejudice includes the very real risk of invoking negative stereotypes and biases against gay, lesbian, and bisexual people.

Although attitudes have shifted, sexual prejudice and heterosexism persist in American society in a way that profoundly affects our justice system. Negative stereotypes about gay and lesbian people include an incorrect link between homosexuality and criminality.109See Simon, supra note 2, at 67–68; Gregory M. Herek, Gender Gaps in Public Opinion About Lesbians and Gay Men, 66 Pub. Opinion Q. 40, 61 (2002); Michael R. Stevenson, Public Policy, Homosexuality, and the Sexual Coercion of Children, J. Psychol. & Hum. Sexuality 1 (2000); Jordan Blair Woods, The Birth of Modern Criminology and Gendered Constructions of Homosexual Criminal Identity, 62 J. Homosexuality 131, 132 (2015). Research indicates that bisexuals are seen as more likely to be unfaithful to their intimate partners and more likely to infect their partner with a sexually transmitted disease.110See Sexual and Gender Prejudice, supra note 9, at 363. One example of the still powerfully looming prejudice against LGBTQ individuals in the United States was President Donald Trump’s selection of Pastor Robert Jeffress to lead the prayer service preceding his January 2017 inauguration.111Niamh McIntyre, Donald Trump’s Inauguration Prayer Service to Be Led by Preacher Who Says Islam Is ‘Evil’ and Gay People ‘Filthy’, Independent (Jan. 20, 2017), http://‌www.independent.co.uk/‌news/‌world/‌americas/‌preacher-pastor-robert-jeffress-islam-evil-gay-people-filthy-trump-inauguration-prayer-service-a7536926.html. With that decision, the then President-elect legitimized and even exalted an individual who has made no secret of his animus towards gay and lesbian people.112Id. Jeffress “has spread false statistics about the prevalence of HIV among LGBT people, who he said live a ‘miserable’ and ‘filthy’ lifestyle, which inevitably leads to depression and alcoholism.”113Id. Jeffress’s presence at the inauguration serves as one barometer of the strength of anti-gay bias in American society that correlates with an apparent recent uptick in anti-gay attitudes among Americans.114See Harris Poll, supra note 8, at 1. In fact, we seem to harbor more widespread overtly heterosexist sentiments as a society than many justice system actors would choose to admit.

In a courtroom setting, defendants who are perceived as gay are less likely to be treated fairly than those perceived as straight, particularly when accused of sexual assault.115Tisha R. A. Wiley & Bette L. Bottoms, Effects of Defendant Sexual Orientation on Jurors’ Perceptions of Child Sexual Assault, 33 L. & Hum. Behav. 46, 47 (2009). Research indicates that “homonegativity affects the guilt rating in sexual assault trials involving homosexuals. Specifically, homosexual males accused of sexually assaulting heterosexual males . . . are found more guilty than heterosexual males who assault heterosexual females . . . .”116Hill, supra note 10, at 102. As one study concludes, “a homosexual male accused of assaulting a heterosexual male will not receive a fair trial.”117Id. at 105 (emphasis added). Even when otherwise admissible, evidence of a defendant’s sexual orientation offered to support or refute the commission of a sexual assault, carries particular risk of unfair prejudice because of the stereotypes and prejudices that place gay and bisexual people at a disadvantage in the justice system. In sexual assault cases, as mentioned above, defendants who are perceived to be gay or bisexual are subject to particular stereotypes about sexual promiscuity and sexual predation. Although exclusion under legal relevance is generally discretionary,118See Fed. R. Evid. 403; Edward J. Imwinkelried, The Meaning of Probative Value and Prejudice in Federal Rule of Evidence 403: Can Rule 403 Be Used to Resurrect the Common Law of Evidence?, 41 Vand. L. Rev. 879, 893 (1988). the risk of unfair prejudice warrants liberal exercise of that discretion.

This evidence arises only in cases where evidence of being gay or bisexual is offered as evidence of guilt, and occasionally, where evidence of being heterosexual is offered as evidence of innocence. Turning back to State v. Ford,119State v. Ford, 926 P.2d 245, 250 (Mont. 1996). imagine for a moment that the victim in that case were a woman. The prosecutor would say, “We have a man who is an admitted bisexual, and that’s exactly what he is charged with doing.” The appellate court decision would say that the defendant’s bisexuality was relevant to show he “fit the profile” of a man who would rape a woman. The arguments are simply absurd in that context.

Some might argue that this gender disparity exists because of the need to contradict a default assumption that a defendant is straight. Evidence of homosexuality, so the argument goes, is necessary because without it most fact-finders will automatically believe that a defendant is heterosexual. However, a diligent prosecutor would not rely on assumption when she had truly probative evidence. Even if most fact-finders would tend to designate a default heterosexual orientation to most defendants, there exists the possibility that some would not. Given that possibility, if the logic holds, a prudent prosecutor would offer evidence that the defendant was heterosexual in order to ensure that the fact-finder drew the desired conclusion. Yet, prosecutors do not offer such evidence in that context.120See supra note 12. That is because there is nothing to gain from it. Even if a juror assumes a defendant is gay, for reasons discussed above, that fact would still carry a bias towards believing that the defendant was more likely to commit a sexual assault, even if the victim is a woman.121See Simon, supra note 2, at 63.

It is fundamentally heterosexist to say that a defendant’s sexual orientation tends to prove sexual assault when that argument is only applied to gay and bisexual people. Even accepting that sexual orientation may be a loose proxy for capacity for sexual desire towards a member of a particular gender (notwithstanding the complications discussed above),122See supra Section II.B. evidence of sexual orientation should operate in the same way, regardless of the sexual orientation of the defendant and gender of the victim. Instead, sexual orientation evidence is applied in a way that creates a double-standard in same-gender sexual assault cases. Being gay is considered probative of guilt and being straight is considered probative of innocence. This unfairly benefits criminal defendants who fit some imagined notion of what it means to be straight, and equally disadvantages those who do not. At its root, this phenomenon serves to police masculinity and punish homosexuality and bisexuality as aberrant and wrong.

This is the fundamental weakness of a legal relevance frame as applied to evidence of a defendant’s sexual orientation in sexual assault cases. Legal relevance analysis relies upon the judge’s ability to accurately assess both the probative value and the risk of unfair prejudice, which means the risk that evidence will be misused, not necessarily just the risk of bias. Misuse can simply mean overvaluing the evidence.123See supra text accompanying 106–07. Given the difficulty in assessing the probative value of evidence of sexual orientation,124See supra Section II.B. the integrity of a legal relevance analysis in this context is vulnerable to corruption, even when judges are acting without malice. That challenge is compounded by the difficulty in facing and accurately measuring the level of anti-gay and anti-bisexual bias present our society.

While overt stereotypes and prejudice are readily identifiable, sexual prejudice can be subconscious and may exist even among those who believe themselves to be supportive of LGBTQ equality. As developmental psychologist Rich Savin-Williams puts it: “Sexual prejudice can be subtle. For example, whereas heterosexual individuals may publicly espouse civil rights for lesbians and gay men as a social group, they may nonetheless experience disgust with same-sex behavior or political activism—expressions of which may ‘leak out’ without their awareness.”125Sexual and Gender Prejudice, supra note 9, at 359.

Thus, judges faced with the legal relevance balancing test must not only measure the elusive probative value of sexual orientation evidence, but also overcome an apparently common tendency to underestimate the risk of unfair prejudice that is created by persisting anti-gay and anti-bisexual biases in American society. While geography and culture may have some effect on those values, even fact-finders in sexually progressive settings are at risk of overestimating the former and underestimating the latter. In fact, because these biases operate subconsciously, those in progressive geographical pockets may be particularly vulnerable to overconfidence in a belief that anti-gay and anti-bisexual bias is a thing of the past.

B. Character Evidence

The doctrine of character evidence developed as an outgrowth from the concept of legal relevance based upon normative values of British and American society.126Roger Park & Tom Lininger, The New Wigmore: A Treatise on Evidence § 1.2 (2018). The widely adopted exclusionary rule for propensity evidence, as captured by FRE 404 is, “[e]vidence of a person’s character or character trait is not admissible to prove that on a particular occasion the person acted in accordance with the character or trait.”127Fed. R. Evid. 404(a)(1). This rule has a long history in common law,128See Jane Harris Aiken, Sexual Character Evidence in Civil Actions: Refining the Propensity Rule, 1997 Wis. L. Rev. 1221, 1225–26 (providing an overview of common law history on the propensity bar and noting it was “the law of the land” as of the eighteenth century in England). rooted in a philosophical commitment to individualism and self-determination—the belief that a person is more than the sum of his traits.129See David P. Leonard, In Defense of the Character Evidence Prohibition: Foundations of the Rule Against Trial by Character, 73 Ind. L.J. 1161, 1199–1200 (1998). Propensity character evidence is considered minimally probative and highly susceptible to overvaluation.130Park & Lininger, supra note 126, § 3.1. Character doctrine evolved in recognition of the fact that character evidence is inherently unfairly prejudicial and should not be left to a case-by-case analysis by individual judges. Instead, a broad rule emerged.

Courts have differed significantly in whether they treat evidence of a sexual assault defendant’s sexual orientation as character evidence.131Compare State v. Schweppe, 237 N.W.2d 609, 615 (Minn. 1975) (stating that evidence of homosexuality “may have tended to impeach [defendant’s] character”), with State v. Ford, 926 P.2d 245 (Mont. 1996) (failing to discuss issue of impermissible character evidence). In one sense, it is too simplistic to say that sexual orientation is a character trait. However, the key question in deciding whether character rules should apply is the proponent’s theory of relevance. When evidence of a defendant’s sexual orientation is offered to show that the defendant has a higher likelihood to engage in same-gender sexual relations and therefore would be more likely to act in conformity with that probability in committing the charged acts, that is the definition of propensity character evidence. In other words, as long as sexual orientation is treated like a character trait, it should be analyzed like one under the rules of evidence.

Looking back at the two exemplar cases discussed in Part I, it is clear that the courts in both Ford and J.O. admitted defendant’s sexual orientation on a character theory of relevance and did so without applying the character rules. This oversight represents another fundamental problem with sexual orientation evidence in sexual assault cases—it is subject to improper use as propensity evidence and courts seem to have trouble recognizing it as such and applying the correct analysis under character rules.132See supra Part I.

In Ford, even as it focused on the notion of what type of person would supposedly commit the crime in question, the majority opinion failed to even mention the issue of impermissible character evidence.133Three Justices joined Justice Trieweiler’s majority opinion which did not discuss character evidence. In a specially concurring opinion, Justice Leaphart argued that the evidence was inadmissible under Montana’s character evidence rule barring propensity evidence, Montana Rule of Evidence 404(a)(1)(a). However, he agreed with the result that defendant’s substantial rights had not been prejudiced by admission of the evidence due to the overwhelming evidence of guilt. Justice Trieweiler filed his own specially concurring opinion only to note that he disagreed with Justice Leaphart’s assertion that sexual orientation constituted character evidence. Ford, 926 P.2d at 250; id. at 252–53 (Leaphart, J., concurring); id. at 253 (Trieweiler, J., concurring). The court’s own assertion that Ford “fit the profile of someone who would commit the act for which he was accused” is a propensity theory—that defendant is a bisexual person and therefore would have acted in conformity with his bisexuality in committing this crime.134For a more complete discussion of propensity character evidence, see infra Section IV.B. The evidence should have been excluded under the general bar on propensity evidence; but the court wholly overlooked the propensity issue in its analysis.

Similarly, the judge in J.O. entirely overlooked the propensity issue.135See J.O. v. O.E., 100 A.3d 478 (D.C. 2014). In saying that the defendant “is not going to approach [the petitioner] for sex,”136Id. at 480. the judge relied upon a propensity theory of relevance, which should have triggered the bar on propensity character evidence in civil cases.137The District of Columbia generally follows FRE 404 as to character evidence. Cf. Fed. R. Evid. 404(a)(1); Drew v. United States, 331 F.2d 85, 89 (D.C. Cir. 1964) (holding that “evidence of one crime is inadmissible to prove disposition to commit crime,” unless relevant to motive, intent, absence of mistake or accident, common scheme or plan, or identity); Rauh v. Coyne, 744 F. Supp. 1181, 1184 (D.D.C. 1990); Smith v. United States, 26 A.3d 248, 260 (D.C. 2011) (“While this jurisdiction has not adopted the Federal Rules of Evidence, this court will look to those rules for guidance. . . . ” (quoting Goon v. Gee Kung Tong, Inc., 544 A.2d 277, 280 n.9 (D.C. 1988)); District of Columbia v. Thompson, 570 A.2d 277, 299 (D.C. 1990), vacated in part on other grounds, 593 A.2d 621 (D.C. 1991) (“[A]s a general rule in civil assault and battery cases, neither party’s character ‘is an issue and cannot be the subject of attack, unless it is first attacked or supported by the adversary, or placed in issue by the nature of the proceeding itself.’” (quoting Phillips v. Mooney, 126 A.2d 305, 308 (D.C. 1956))). Yet the judge admitted the testimony into evidence and relied almost exclusively upon it in his ruling.138See J.O., 100 A.3d at 481–82. Both the logical conflation and the refusal to apply applicable legal standards demonstrate that the court’s reasoning was compromised by prejudice. Although the bar on propensity character evidence was a central argument in the appellant’s brief,139Opening Brief of Appellant at 17, J.O. v. O.E., 100 A.3d 478 (D.C. 2014) (No. 12-FM-1949). the appellate court skirted the issue of propensity evidence in a vague and equivocal footnote.140The court’s analysis was preceded by the caveat, “even assuming [defendant’s testimony that he was not gay] was admissible” and a footnote which briefly mentioned character evidence. J.O., 100 A.3d 478 at 481–82 & n.13. The footnote stated that evidence of a character trait is not admissible to prove that a person acted in conformity with that trait and cited District of Columbia v. Thompson for the exception that evidence of peacefulness or aggression to prove which party was the first aggressor is admissible when mutual assault is alleged in a civil assault case in the District of Columbia. Id. at 482 n.13 (citing District of Columbia v. Thompson, 570 A.2d 277, 299 (D.C. 1990)). However, Thompson was inapposite because, in J.O., there was no allegation of mutual assault and the character trait in question was not peacefulness or aggression. This represented the only mention of character evidence in the opinion.

When evidence of a defendant’s sexual orientation is offered on a propensity theory of relevance, the court ought to apply a character evidence analysis. Exclusion under the propensity bar is generally mandatory, which eliminates the risks associated with discretionary exclusion in a legal relevance analysis. However, this is not the end of the analysis. Propensity doctrine has a number of exceptions that must be navigated before it poses a real solution to the fundamental problem confronted by this Article. In addition, even if a piece of evidence survives a character analysis, it still must be subject to a legal relevance balancing test.

The bar on character evidence has several key exceptions, including: (1) witness impeachment, (2) prior sex crimes, (3) the so-called MIMIC exceptions, and (4) the mercy rule. First, the witness impeachment exception allows specific forms of character evidence to be admitted for the purpose of attacking a testifying witness’s character for truthfulness.141Fed. R. Evid. 607–09. Sexual orientation is wholly unrelated to a person’s character for truthfulness, therefore this exception is inapplicable to affirmative evidence about a witness’s sexual orientation. As such, the following discussion focuses on the latter three exceptions.

The prior sex crimes exception, also called sexual character doctrine, codified in the Federal Rules of Evidence at 413, 414, and 415, allows evidence that a defendant committed certain sex offenses to be admitted in sex crime cases, both criminal and civil. The basic rule says, “[i]n a criminal case in which a defendant is accused of a sexual assault, the court may admit evidence that the defendant committed any other sexual assault.”142Fed. R. Evid. 413(a). There is a parallel rule for child molestation cases,143Fed. R. Evid. 414. and one that expands the exception from criminal to civil cases.144Fed. R. Evid. 415. Notwithstanding its critics,145E.g., Sara Sun Beale, Prior Similar Acts in Prosecutions for Rape and Child Sex Abuse, 4 Crim. L.F. 307 (1993); David P. Bryden & Roger C. Park, “Other Crimes” Evidence in Sex Offense Cases, 78 Minn. L. Rev. 529 (1994); Edward J. Imwinkelried, A Small Contribution to the Debate over the Proposed Legislation Abolishing the Character Evidence Prohibition in Sex Offense Prosecutions, 44 Syracuse L. Rev. 1125 (1993); Mark A. Sheft, Federal Rule of Evidence 413: A Dangerous New Frontier, 33 Am. Crim. L. Rev. 57 (1995). sexual character doctrine represents a normative choice that arose from a recognition of the unique nature of sex crimes.146People v. Britt, 128 Cal. Rptr. 2d 290, 293–95 (Cal. Ct. App. 2002); Tamara Larsen, Sexual Violence Is Unique: Why Evidence of Other Crimes Should Be Admissible in Sexual Assault and Child Molestation Cases, 29 Hamline L. Rev. 177 (2006). It functions to except evidence of prior sex crimes from the general character bar. As applied to the issue this Article confronts, sexual character evidence would allow admission of a defendant’s prior commission of a sexual assault in a sex crime prosecution or civil adjudication, but not prior instances of consensual sexual behavior. Notably, there is no restriction as to similarity between the charged acts and the prior offenses147Fed. R. Evid. 413–15.; therefore prior sex crimes would be admissible regardless of the gender of the defendant or victim. Although sexual orientation may be (correctly or incorrectly) inferred from the gender of the defendant and prior victim, that is not the main function of the evidence. When a prior sex crime is one against a member of the defendant’s same gender, that evidence is admissible to prove a specific propensity for sexual violence, the probative value of which substantially overshadows any collateral inference about a defendant’s sexual orientation. This is the key distinction between prior sex crimes and all other forms of sexual orientation evidence.

The so-called MIMIC exception,148MIMIC is a mnemonic acronym that stands for motive, intent, (absence of) mistake, identity, or common scheme or plan. People v. Rojas, 760 N.E.2d 1265, 1268 n.3 (N.Y. 2001) (“As ‘known to generations of bar review students,’ these five categories produced the ‘MIMIC rule.’” (quoting Michael M. Martin, Daniel J. Capra & Faust F. Rossi, New York Evidence Handbook 239 n.99 (1997))).. rooted in the common law and codified at Federal Rule of Evidence 404(b)(2), says that character evidence in the form of prior acts may be admissible if offered on a non-propensity theory of relevance, “such as . . . motive, opportunity, intent, preparation, plan, knowledge, identity, absence of mistake, or lack of accident.”149Fed. R. Evid. 404(b)(2). This is not so much an exception as a clarification that evidence of prior acts, even when they might implicate character, may be admissible if probative of a material fact at issue other than character. Both the mnemonic acronym MIMIC and the more expansive language of 404(b)(2) offer non-exhaustive lists of examples of non-character theories of relevance. The rule itself, again, is quite simple: “[character] evidence may be admissible for another purpose . . . .”150Id. In other words, evidence that is probative of character may be admissible on a non-propensity theory of relevance. Motive, intent, etc. are merely examples of non-propensity theories of relevance. For the purposes of later discussion in this Article, it is critical to understand that although character evidence may consist of prior acts, reputation, or opinion, the MIMIC principle specifically applies to prior acts.151See id.; Fed. R. Evid. 405. The MIMIC principle poses a challenge to courts weighing the admissibility of sexual orientation evidence because a theory of motive or intent can be easily confused with propensity. This will be discussed further in Section IV.B.152See infra Section IV.B.

Finally, the mercy rule, codified in the federal rules at 404(a)(2), allows a criminal defendant to put on evidence of his or her “pertinent [character] trait,” at his or her discretion.153Fed. R. Evid. 404(a)(2). This rule arises from a desire to correct the imbalance of power between the state and the criminal accused as well as a policy preference towards allowing criminal defendants every opportunity to defend their liberty.154Fed. R. Evid. 404 advisory committee notes to 2006 amendment (“In criminal cases, the so-called ‘mercy rule’ permits a criminal defendant to introduce evidence of pertinent character traits of the defendant and the victim. But that is because the accused, whose liberty is at stake, may need ‘a counterweight against the strong investigative and prosecutorial resources of the government.’” (citing C. Mueller & L. Kirkpatrick, Evidence: Practice Under the Rules 264–65 (2d ed. 1999); Richard Uviller, Evidence of Character to Prove Conduct: Illusion, Illogic, and Injustice in the Courtroom, 130 U. Pa. L. Rev. 845, 855 (1982))). Once a defendant raises a character trait, the prosecution is then entitled to present evidence to rebut defense evidence on that specific character trait.155Fed. R. Evid. 404(a)(2).

Generally, the mercy rule is used to allow a defendant to call a character witness to testify that he or she is a good person and would never commit the crime that has been alleged. However, the mercy rule does not necessarily restrict a defendant to that common script.156Id. The mercy rule would allow a defendant, accused of sexual assault, to raise his or her sexual orientation on a propensity theory (i.e., to support the proposition that defendant would not be inclined to desire a member of the same gender and therefore would not commit a desire-motivated sexual assault). Once the defendant placed his or her sexual orientation at issue, by testifying “I’m not gay,” for example, then the prosecution would be free to offer evidence showing that defendant is gay. Had J.O. v. O.E. been a criminal prosecution, it might have looked just like this. However, in civil cases like J.O., which was a civil protection order adjudication, the mercy rule does not apply.157Another aspect of the mercy rule allows a criminal defendant to attack the victim’s “pertinent [character] trait,” subject to the limitations of the rape shield rule. Fed. R. Evid. 404(a)(2)(B); Fed. R. Evid. 412. There has been much written about the subject of rape shield, which is beyond the scope of this Article. See sources cited supra note 19. However, courts seem to have no trouble excluding evidence of victim sexual orientation under the rape shield rule. See, e.g., People v. Murphy, 919 P.2d 191, 195 (Colo. 1996); Minter v. Commonwealth, 415 S.W.3d 614, 619 (Ky. 2013); State v. Boutchiche, No. E2007-00473-CCA-R3-CD, 2009 WL 102949 (Tenn. Crim. App. Jan. 12, 2009).

C. When the Risk of Unfair Prejudice Is Clear: Analogies

Although judges often underestimate the risk of unfair prejudice and admit evidence of defendant sexual orientation in adult sexual assault cases,158See cases cited supra note 11 and accompanying text. in analogous types of cases, courts are far more likely to exclude such evidence.159See infra Sections III.C.1–2. The following Sections explore two sets of cases in which judges tend to recognize that the risk of unfair prejudice vastly outweighs any probative value and exclude evidence of defendant (or sexual assailant) sexual orientation: “gay panic” homicide cases and child sex abuse cases. These cases are similar to adult sexual assault cases in many ways, presenting identical evidentiary issues. However, courts have treated those evidentiary issues quite differently.

In each of these comparator sets of cases, harmful anti-gay and anti-bisexual stereotypes are more explicit and egregious, and thus easier for judges to recognize. As a result, courts have been more careful with evidentiary rulings in these cases than adult same-gender sexual assault cases. This heightened caution is appropriate. However, when peeled apart, the evidentiary issues in these comparator sets turn on the same questions as adult same-sex sexual assault cases and the same analysis should apply in each, even if the risk of unfair prejudice is quantified differently. Although, or perhaps because, the harm may be harder for judges to recognize in adult sexual assault cases, similar caution should be exercised in those cases. These comparator sets of cases show that judges are, when acting cautiously and in light of anti-gay and anti-bisexual stereotypes, able to use existing evidence law doctrine to properly handle sexual orientation evidence in sexual assault claims.

1. “Gay Panic” Cases

The first instructive comparison is found in a subset of “gay panic” cases, in which a defendant claims imperfect self-defense against a sexual assault by a victim and raises evidentiary issues similar to those examined by this Article.160See Pei-Lin Chen, supra note 22, at 210–11. Infamously raised during the Matthew Shephard and Lawrence King murder trials, the “gay panic” defense is invoked when, as Professor Cynthia Lee explains, “male defendants charged with murdering gay men have sought mitigation or exoneration by claiming gay panic, either as a manifestation of mental disease or defect or as support for a claim of provocation or self-defense.”161Cynthia Lee, The Gay Panic Defense, 42 U.C. Davis L. Rev. 471, 489 (2008). These arguments may arise in an assault or homicide case in which a defendant claims to have assaulted or killed a same-gender victim in response to either a sexual overture or sexual assault by that victim.162Id. at 474; see also Nicolas, supra note 64, at 809. The theoretical basis for this defense is that the prospect of a sexual advance or assault by a same-gender victim was so upsetting as to provoke a temporary loss of faculties and trigger a violent, even fatal, response.163Lee, supra note 161, at 483 (describing the historical origins of “homosexual panic,” first conceived of as a mental disorder). As Lee has explains, this defense can hang on the “doctrinal hook” of excuse, justification, insanity, diminished capacity, or provocation.164Id. at 494. Each of these arguments rests on deeply heterosexist foundations and privileges the absolute defense of straightness and heteronormative masculinity above the lives and safety of gay and transgender165While gay panic and trans panic tropes bear some commonalities, the violence faced by trans people is distinct and merits significant attention from academics, advocates, and lawmakers. people. Scholars have been highly critical of “gay panic” defenses166E.g., Robert B. Mison, Homophobia in Manslaughter: The Homosexual Advance as Insufficient Provocation, 80 Calif. L. Rev. 133 (1992); Pei-Lin Chen, supra note 22, at 210–11; Kara S. Suffredini, Note, Pride and Prejudice: The Homosexual Panic Defense, 21 B.C. Third World L.J. 279, 302 (2001).; and advocates are currently engaged in a multi-state effort to ban “gay and trans panic” defenses from being raised in court.167Jaclyn Cosgrove, LGBTQ Advocates Welcome End of ‘Gay Panic’ Defense in Illinois, Look to Other States, LA Times (Dec. 29, 2017, 11:35 AM), http://‌www.latimes.com/‌nation/‌la-na-lgbtq-gay-panic-defense-20171228-story.html; Associated Press, Illinois to Become Second State to Ban ‘Gay Panic Defense’, NBC News (Dec. 29, 2017, 2:15 PM), http://‌www.nbcnews.com/‌feature/‌nbc-out/‌illinois-become-second-state-ban-gay-panic-defense-n833441. The American Bar Association has taken a position in favor of banning gay and trans panic defenses.168ABA House of Delegates, Res. 113A (2013), http://‌lgbtbar.org/‌wp-content/‌uploads/‌2014/‌02/‌Gay-and-Trans-Panic-Defenses-Resolution.pdf. However, the gay panic defense has only been banned in two states and is therefore theoretically permissible in most of the country.169S.A.B. 2501, 2014 Leg., Reg. Sess. (Cal. 2014) (amending California’s manslaughter code, Cal. Penal Code § 192(f)(1) to state that provocation is not objectively reasonable if based upon victim sexual orientation or gender); S.A.B. 100-0460, 2017 Gen. Assemb., Reg. Sess. (Ill. 2017) (amending section 9-2 to state that serious provocation cannot include discovery of victim sexual orientation).

Courts’ handling of gay panic defense claims has been far from perfect, but a general trend to exclude evidence of the victim’s sexual orientation has emerged, providing an example of how the evidentiary issues can be handled, when the risk of unfair prejudice is more fully appreciated.170People v. Miller, 981 P.2d 654, 658 (Colo. App. 1998) (holding that evidence of specific acts offered to prove homicide victim’s sexual orientation was inadmissible character evidence when offered to show that victim made an unwanted sexual advance towards defendant); Blair v. State, 543 S.E.2d 685, 687 (Ga. 2001) (holding that a trial court did not abuse its discretion in excluding evidence of a victim’s alleged solicitation of sex from a witness); State v. Divers, 889 So. 2d 335, 347 (La. Ct. App. 2004) (holding that trial court did not violate murder defendant’s rights by preventing him from presenting evidence about victim’s sexual orientation); State v. Laws, 481 S.E.2d 641, 647 (N.C. 1997) (holding that evidence of homicide victim’s homosexuality was inadmissible and had little tendency to show that the victim was the aggressor where defendant claimed killing in response to victim’s “homosexual advance.”); cf. People v. Covich, 241 A.D.2d 932, 932 (N.Y. App. Div. 1997) (holding that New York’s rape shield law precluded defense witness testimony as to his same-sex relationship with victim and victim’s sex with male prostitutes); State v. Bell, 805 P.2d 815, 817 (Wash. Ct. App. 1991) (holding that victim’s “homosexual reputation” was inadmissible as both irrelevant to self-defense and unfairly prejudicial). But see State v. Lowe, 505 N.W.2d 662, 669 (Neb. 1993) (holding that evidence of murder victim being gay was admissible in support of defendant’s self-defense claim). Even outside the “gay panic” homicide context, courts have held that evidence of sexual orientation is inadmissible to support an inference of sexual contact. See, e.g., State v. Anderson, 358 So. 2d 276, 277–78 (La. 1978) (holding no abuse of discretion when trial court disallowed defense questioning of kidnapping victim regarding past “homosexual activities” to support theory that male victim consented to go with male defendant to engage in sex). For the purposes of this Article, the most useful inquiry is into the subset of primarily homicide cases in which a defendant attempts to raise the victim’s sexual orientation to support a claim that the victim attempted to sexually assault the defendant (rather than make a non-assaultive overture). In such cases, over the past several decades, judges have tended to exclude evidence of the victim’s sexual orientation based on a variety, and sometimes a combination, of evidentiary doctrines.171See cases cited supra note 170. From an analytic perspective, these cases confront the same evidentiary question as same-gender sexual assault cases discussed in this Article: Is an alleged sexual assailant’s (homicide victim’s) sexual orientation admissible to prove that the alleged sexual assailant (homicide victim) committed or attempted sexual assault against the sexual assault victim (homicide defendant)? Yet courts generally answer this question quite differently than in same-gender sexual assault cases.

In this subset of gay panic cases, courts have excluded homicide victims’ sexual orientation as logically irrelevant, legally irrelevant, or impermissible character evidence.172See cases cited supra note 170. This trend seems to be driven by a recognition that anti-gay stereotypes and beliefs undergird the gay panic defense in particularly harmful ways. Yet, instead of banning defendants from raising the gay panic defense as an improper self-defense claim, many courts have instead focused on excluding evidence of victim sexual orientation under the rules of evidence. In doing so, judges have successfully deployed the rules of evidence, mindful of anti-gay stereotypes inherent in relevance arguments about sexual orientation evidence, with logic that could also apply in other sexual assault cases.

The gay panic defense generally presumes that a same-gender sexual overture or assault is so justifiably repulsive to a straight man that he should not be held fully legally responsible for responding with lethal violence.173See Lee, supra note 161; see also Nicolas, supra note 64, at 809–819. In other words, it explicitly devalues the lives of gay people. This notion is antithetical to the fundamental principles of justice, and perhaps that is why many courts prefer to exclude sexual orientation of the victim (the alleged sexual assailant) in these cases.174A recent wave of advocacy seeks to ban the gay panic defense by prohibiting it from being raised at trial. Cosgrove, supra note 167. However, whether the defense is palatable or moral is not part of the legal relevance analysis as to a particular piece of evidence. An unpalatable claim of gay panic self-defense does not necessarily increase the risk of unfair prejudice generated by a particular piece of evidence. Instead, the more appropriate remedy would be for courts (or legislatures) to bar gay panic as an improper self-defense argument.

So, while gay panic claims that rely on the presumed validity of violent defense of heterosexual masculine integrity are deeply problematic, such arguments do not create an evidentiary distinction between same-gender sexual assault cases and gay panic defense cases. Interestingly, when excluding victim sexual orientation evidence in sexual assault cases, courts do not cite these inherent problems with the gay panic defense. Instead, judges conduct a rigorous application of multiple evidentiary doctrines, ultimately tending to exclude the evidence. These analyses rest on reasoning that applies equally to sexual assault cases. For example, several courts have expressly noted the heterosexism inherent in arguments made by the proponents of sexual orientation evidence in gay panic cases. As the Supreme Court of North Carolina said in State v. Laws, “[a] victim’s homosexuality has no more tendency to prove that he would be likely to sexually assault a male than would a victim’s heterosexuality show that he would be likely to sexually assault a female.”175481 S.E.2d at 647. This conclusion should apply as much to gay panic homicide cases as conventional sexual assault cases.

2. Child Sex Abuse Cases

Another context in which questions of sexual orientation evidence arise is child sex abuse claims. As in gay panic cases, courts have more successfully recognized the risk of harm posed by anti-gay stereotypes and prejudice in child sex abuse cases. As a result, a more rational application of the rules of evidence emerges. Courts generally exclude evidence of a defendant’s sexual orientation in child sex abuse cases, recognizing the low probative value and high risk of unfair prejudice.176See Sias v. State, 416 So. 2d 1213, 1217 (Fla. Dist. Ct. App. 1982) (holding admission of homosexuality was irrelevant and highly prejudicial); Phillips v. State, 350 So. 2d 837, 838 (Fla. Dist. Ct. App. 1977) (holding that use of defendant’s homosexuality to prove bad character is reversible error); State v. Young, No. 102,121, 2013 WL 6839328, at *19 (Kan. Ct. App. Dec. 27, 2013) (holding defendant’s literary collection by and about homosexuals does not indicate anything about defendant’s orientation, but creates high susceptibility for jury to make this assumption and attach any unfounded stereotypes to the defendant); State v. Tizard, 897 S.W.2d 732, 745 (Tenn. Crim. App. 1994) (finding a substantial risk that jury would at least partially rely on defendant’s sexual orientation to determine guilt). But see State v. Ditzler, No. 00CA007604, 2001 WL 298233, at *8–9 (Ohio Ct. App. Mar. 28, 2001) (holding that defendant’s homosexuality was highly probative when the alleged offense was male-on-male rape). In doing so, judges have issued passionate rejections of the “odious and unfounded stereotype that homosexuals are more likely to be pedophiles . . . .”177Young, 2013 WL 6839328, at *19; see also Sias, 416 So. 2d at 1217. A Kansas appeals court applied this logic in State v. Young, holding that a defendant’s book collection, even when it included gay-themed titles, was inadmissible:

Indulging the fatuous notion that persons with literary or cinematic works by or about gays or lesbians more likely than not share that sexual orientation has little or nothing to do with this case. Only by taking that assumption and applying to it the odious and unfounded stereotype that homosexuals are more likely to be pedophiles than heterosexuals can those books have anything to do with this case. This court and others have rejected that argument and the false premise upon which it is based as wholly improper.178Young, 2013 WL 6839328, at *19.

In the context of child sex abuse cases, some judges have gone a step further, not only rejecting the false link between homosexuality and pedophilia, but also explicitly surfacing and rejecting the stereotype that gay, lesbian, and bisexual people are more likely to commit sex crimes against any members of the same gender, regardless of the victim’s age. In upholding the exclusion of a police officer’s testimony that a male defendant was gay, a Florida court of appeals outlined and debunked two separate logical inferences: first, that a gay defendant was more likely to have engaged in sexual assault against a member of the same gender, and second, that a gay defendant was more likely to have engaged in child sexual abuse:

While the State suggests that the relevance of this testimony is that only one who is a homosexual would commit a homosexual battery, the suggestion is not well taken. First, the homosexual populace is certainly large enough as to make the relevance of this testimony so tenuous as to be almost non-existent. Second, there is absolutely no showing that homosexuals as a group are disposed to engage in pederasty.179Sias, 416 So. 2d at 1217.

It would have been sufficient to guarantee a correct result in the case if the court had rested its argument solely on the improper inference that gay men are predisposed to pedophilia. But instead, the court chose to take the additional step of attacking the inference that only a gay man would commit sexual assault against a male victim.

Similarly, in State v. Tizard, a Tennessee court conducted a measured analysis of evidence of the defendant’s sexual orientation, using logic that also would apply in an adult sexual assault case.180897 S.W.2d at 743–45. Physician Gary Tizard was convicted of two counts of sexual battery by fraud arising out of conduct towards a seventeen-year-old boy he was treating.181Id. at 735–36. At trial, the state presented sexually explicit materials found in defendant’s possession.182Id. at 735. The evidence included a pamphlet called “The Art of Marathon Masturbation,” which depicted men engaged in masturbation, and a pornographic video showing sex between two men.183Id. at 738. Defense counsel sought to exclude the evidence, arguing it was irrelevant and unfairly prejudicial.184Id. at 743. The trial court chose to exclude the contents of the videos and pamphlet themselves, but allowed testimony about the contents, stating it was relevant and admissible “to show any proclivity of the Defendant to commit the offense for which he is charged.”185Id. (internal quotation marks omitted). The State also called a detective who testified that defendant stated his two roommates were “possibly bisexual.”186Id. at 744. If there was any doubt about how the prosecution expected that evidence to be used, in closing, the State argued:

There was a proverb, ladies and gentlemen, “As a man thinketh, so is he.” As a man thinketh, the art of marathon masturbation, so is he. As a man thinketh, in the back of his car a sexually explicit video depicting sexual acts of men, so is he. As a man thinketh, videos of sexually explicit acts between men in the recreation room and the master bedroom, so is he.187Id. at 745.

Defendant appealed his conviction, arguing that the video and pamphlet were improperly admitted.188Id. at 735. On appeal, the State of Tennessee essentially conceded that the evidence should not have been admitted as propensity character evidence, calling the “proclivity” language an “unfortunate mischaracterization.”189Id. at 743. However, the State argued that the evidence was admissible to prove the intent element of sexual assault in the Tennessee statute, invoking the MIMIC principle discussed in Section III.B. above.190Id.. The Tennessee Court of Appeals reversed the conviction, specifying not only that the evidence constituted improper propensity evidence, but also that it “was not rationally related to the issue of the defendant’s criminal intent and should have been excluded.”191Id. at 744.

In reaching that conclusion, the Tennessee Court of Appeals used a methodical and logically sound approach that should serve, for the most part, as an example to any judge faced with this type of evidence. The court began by correctly concluding that the evidence should be inadmissible as to a propensity theory of relevance as improper character evidence.192Id. at 745. The court then outlined the next steps, to assess logical relevance and conduct a legal relevance balancing test: “The question then becomes what relevance does [the evidence] have to show that [the defendant had the requisite intent].”193Id. at 744. “If any probative value on the issue of intent exists other than through showing propensity, then the inquiry is whether that value is outweighed by the danger of unfair prejudice.”194Id.

The court highlighted the logical flaws and inherent heterosexism embedded in the State’s argument by analogizing the evidence in question with heterosexual pornography, noting that the State’s argument was the same as “contending that the defendant’s possession of heterosexually explicit videotapes and a book regarding female masturbation would tend to prove that the defendant committed a sexual battery upon a female patient.”195Id. Turning first to the question of logical relevance on the issue of intent, the court examined whether the evidence could clear the threshold of having some probative value to a material fact at issue. Central to the analysis was a separation of two logical inferences: (1) whether the evidence tended to prove the defendant was gay, and (2) whether being gay tended to prove that defendant possessed the requisite criminal intent.196Id. The court concluded that the first inference could possibly be made, but said “the hurdle lies in attempting to translate this into further legitimate inferences which are sufficiently relevant to an intent to commit a sexual battery upon the victim.”197Id. (citation omitted). The court then explained that the logical link between the evidence and the second inference (intent) was too great and too dependent upon the questionable first inference (homosexuality), stating that the “hurdle” of linking the two inferences was “insurmountable.”198Id. Ultimately, the court concluded that the evidence was “not rationally related” to the relevant inference.199Id.

The court also astutely noted that the intent theory of relevance was indistinct from an impermissible propensity theory. In fact, the prosecution’s purported intent theory of relevance was a propensity theory. As the court said, “we do not tolerate the use of evidence merely showing a propensity or character trait in order to prove that an accused acted in conformity with the propensity or trait so as to commit a given offense or possessed the particular criminal intent which is required for that offense.”200Id. at 744–45 (emphasis added); cf., e.g., United States v. Henry, 848 F.3d 1, 8, 9 (1st Cir. 2017) (holding lower court did not abuse its discretion in admitting the defendant’s prior drug conviction, but noting that the “conclusion [was] compelled by the combination of our deferential standard of review and our precedent,” and that “in many cases, impermissible propensity reasoning lurks as one of the links in the logical chain of relevance”). The pornographic materials were only relevant to intent in that they tended to show the defendant was gay and therefore would have acted in conformity with that trait by intentionally seeking sexual gratification with a member of the same gender. Therefore, the evidence should be excluded under the propensity bar.

Thus, the Tennessee Court of Appeals conducted a thorough analysis of the evidence and found it was improperly admitted at trial. The logic of their approach would be just as sound if the victim were an adult. Certainly, the risk of unfair prejudice related to a defendant’s sexual orientation is higher in child sex abuse cases than in adult sexual assault cases201Cf. State v. Young, No. 102,121, 2013 WL 6839328, at *19 (Kan. Ct. App. Dec. 27, 2013); Sias v. State, 416 So. 2d 1213, 1217 (Fla. Dist. Ct. App. 1982).; and the bigoted notion of a link between homosexuality and pedophilia belongs nowhere in our judicial system. However, the court in Tizard was looking at much more than just a legal relevance balancing test. Crucially, the character analysis is entirely independent of the risk of prejudicial effect.202See Tizard, 897 S.W.2d at 744. Overall, this case demonstrates a rational and non-discriminatory approach to managing evidence of defendant sexual orientation that is applicable in adult sexual assault cases.

IV. Old Framework, New Approach

In this Part, I offer some guidelines for applying existing evidence law principles to evidence of a defendant’s sexual orientation in sexual assault cases. The key component of my approach is a precise and deliberate procession through legal relevance, character, and logical relevance rules. Adherence to this approach should result in the exclusion of defendant sexual orientation evidence in most circumstances. In particular, this proposal suggests that courts exercise their discretion to exclude evidence of a defendant’s sexual orientation as legal relevance, absent a specific articulation of probative value that outweighs the prejudicial effect inherent in this type of evidence. Through the guidelines below, courts can mitigate heterosexism and ensure fairer results, particularly in same-gender sexual assault cases.

A. Relevance Revisited

As with all evidence, the first question regarding admissibility of defendant sexual orientation evidence in sexual assault cases is one of logical relevance. Unless the proponent specifically articulates another theory of relevance (such as corroboration of victim testimony), a court should assume that this evidence is offered on a propensity theory.203See supra Section III.C.2. That is, the evidence is offered because it tends to show that the defendant would be inclined towards assaulting a member of the same gender. That probative link can be reached through either of two separate and critically different inferences. The first is that being gay makes a defendant more likely to commit sexual assault. The second is that being gay makes a male defendant who commits a sexual assault more likely to select a male victim.

The first inference—that being gay makes a defendant more likely to commit sexual assault—is simply false.204A. Nicholas Groth & H. Jean Birnbaum, Adult Sexual Orientation and Attraction to Underage Persons, 7 Archives Sexual Behav. 175 (1978). It is based on sexual prejudices and negative stereotypes about gay, lesbian, and bisexual people that have no place in our justice system.205See supra notes 109–10 and accompanying text. The danger of this inference is that it lurks behind any other use of sexual orientation evidence, which might be improperly used by a fact-finder to support this inference, either consciously or unconsciously. I will revisit this issue in Section C below. The second inference—that being gay makes a person who commits a sexual assault more likely to select a male victim—is deeply complicated and attenuated, but ultimately conceivable, for reasons discussed in Part II of this Article, but only insofar as sexual orientation is an imperfect proxy for desire and only in cases where desire forms part of the theory of the case. However, the narrowness of this inference cannot be overstated. It is little more meaningful than an inference that being a classical music fan makes a person who steals a CD more likely to steal a classical album. Depending on the factual circumstances of a case, evidence that purports to be relevant via a victim selection inference may not survive a logical relevance threshold test. In cases where desire is asserted as a motive, evidence of the defendant’s sexual orientation may be logically relevant. However, that evidence would be subject to additional scrutiny under the character evidence rules and a final legal relevance balancing test.

Within the logical relevance inquiry, judges must also closely scrutinize the evidence itself to determine, even if the defendant’s sexual orientation is relevant, whether the evidence is in fact probative of the defendant’s sexual orientation. The court must determine if the evidence offered really does tend to prove that the defendant is gay. Evidence that someone possessed a book featuring storylines about gay men does not survive a logical relevance test, as the court in State v. Young deftly explained.206Supra text accompanying note 178 (quoting Young, 2013 WL 6839328, at *19). On the other hand, evidence that a defendant self-identifies as gay or that he has had past sexual relationships with men may be relevant, but only to victim selection on a desire theory. However, that assessment will be highly fact-specific. If the court determines that the evidence is relevant, the analysis then continues to the character rules.

B. Propensity by Any Other Name

Character doctrine, when properly applied, will exclude most defendant sexual orientation evidence; but courts must recognize sexual orientation evidence as propensity evidence and be highly skeptical of evidence offered under propensity exceptions. As discussed in Section II.B.1 above, sexual orientation is a label given to identity, desire, and behavior, which often, but not always, overlap.207Supra Section II.B.1. For example, a defendant’s own testimony that he identifies as gay may make it more likely that he would feel sexual desire towards a member of the same gender,208Supra Section II.B.1. which is precisely an impermissible propensity theory.209Supra text accompanying note 154. The great challenge of the character rules as applied to sexual orientation is posed by the exceptions. Judges must be vigilant in identifying propensity theories of relevance masquerading as other theories, such as motive or intent, as in State v. Tizard. Despite this potential pitfall, when properly applied, the character rules will result in the exclusion of sexual orientation evidence in nearly all cases, except when offered by a criminal defendant under the mercy rule. In addition, even the evidence that does survive under a character exception will be subject to final scrutiny under a legal relevance balancing test discussed below.

As applied to sexual orientation, like many other types of character evidence, the so-called MIMIC principle threatens to be the exception that swallows the rule against propensity evidence. Courts must therefore use heightened caution when examining defendant sexual orientation evidence purportedly offered on a non-propensity theory of relevance such as motive or intent. As discussed in Section III.B, the MIMIC principle is not an exception that makes propensity evidence admissible, but a set of examples of admissible non-propensity theories of relevance that may make evidence that implicates a person’s character relevant for some reason other than a character inference.210Supra Section III.B. Still, in many jurisdictions, admissibility for intent in particular has expanded well beyond requiring a non-propensity theory of relevance.211Cf., e.g., United States v. Henry, 848 F.3d 1, 8–10 (1st Cir. 2017) (holding there is no abuse of discretion in admitting evidence of the defendant’s prior drug conviction, based upon judicial precedent); Rafferty v. State, 16 S.W. 728, 730 (Tenn. 1891) (holding that the defendant’s prior insurance fraud was admissible in order to show likelihood of committing the charged acts in an insurance fraud prosecution, based upon judicial precedent). But even in its bloated form, the law on admissibility of intent evidence ought to be limited to prior criminal offenses, not simply the fact of someone’s sexual orientation.212See Fed. R. Evid. 413–15; cf. Henry, 848 F.3d at 8–10; Rafferty, 16 S.W. at 730. Thus, evidence of sexual orientation alone, without evidence of prior sex crimes, would not be admissible as intent evidence.

When evidence that implicates the sexual orientation of a defendant is offered in a sexual assault case on a non-propensity theory of relevance, the judge must carefully parse the logical chain linking the evidence to a material fact at issue in order to identify and exclude propensity evidence hidden within a series of inferences that ends with motive, intent, etc. For example, if a male defendant’s history of having sex with men is offered to show intent to sexually assault a male victim, separating the inferences within that logical chain reveals that the intent theory still relies upon a propensity inference. As demonstrated in the discussion of Tizard in Section III.C.2 above, evidence of propensity to form intent is still improper propensity evidence.213Supra Section III.C.2. Similarly, sexual orientation evidence offered to show motive (e.g., same-gender sexual desire) is still improper propensity evidence because it relies upon the inference that someone who is gay has a propensity to feel sexual desire towards other men. Ending a logical chain with a motive or intent inference does not erase a reliance on a propensity inference within that chain, which makes the evidence inadmissible.

Character doctrine is a source of much confusion among law students, jurists, and litigators alike. It is an area of law widely recognized as illogically founded and inconsistently applied.214See, e.g., Phillip W. Broadhead, Why Bias Is Never Collateral II: Necessary Limitations on Attempts to Rehabilitate Impeached Witnesses in Criminal Cases, 34 Am. J. Trial Advoc. 239, 253–55 (2010). The rules of character evidence were famously called a “grotesque structure” of “misshapen stone[s]” by Justice Robert Jackson in Michelson v. United States.215Michelson v. United States, 335 U.S. 469, 486 (1948). However, Justice Jackson went on to say that character doctrine had “proved a workable even if clumsy system when moderated by discretionary controls in the hands of a wise and strong trial court.”216Id. at 486. The confusing nature of the character rules is a potential drawback of my proposal to reinvigorate the use of propensity doctrine as applied to sexual orientation evidence. However, given the complexity of the issue and the competing interests at stake, a simple solution is not a realistic goal. As Justice Jackson also said, “[t]o pull one misshapen stone out of the grotesque structure is more likely simply to upset its present balance between adverse interests than to establish a rational edifice.”217Id. Thus, my proposal is not to dismantle the structure of the character rules, nor an attempt to pile new stones onto it. Instead, I propose that courts apply the existing rules methodically and in a manner that is scrupulously conscious of both the logical links underlying evidentiary offerings and the pernicious anti-gay and anti-bisexual biases that pervade our justice system.

One facet of the complexity of character doctrine that is valuable as applied to sexual orientation evidence is the distinction between civil and criminal cases regarding the mercy rule. As discussed in Section III.B, the mercy rule suspends the bar on propensity evidence and allows criminal defendants to put on evidence of their character; but the form of that evidence is limited to reputation in many jurisdictions.218Supra Section III.B. As applied to sexual orientation evidence in sexual assault cases, the mercy rule would allow defendants to put on evidence of their sexual orientation, generally only in the form of reputation evidence. Once the door is opened, the prosecution can then test knowledge of the character witness by inquiring about specific acts and put on its own reputation witness to rebut the defendant’s evidence as to his or her sexual orientation.219Cf. supra Section III.B. This compromise allows the defendant, whose liberty interests are at stake in a contest against the formidable power of the state, to make an individual determination about whether to open the door to his or her sexual orientation in a specific case. At the same time, in a civil case, such as the civil protection order case J.O. v. O.E., no such exception is made, and the civil respondent could not invoke an “I’m not gay” defense, nor could the petitioner put on evidence that the defendant was gay.

It is important to highlight the difference between sexual orientation evidence and prior sex crimes evidence, which is specifically admissible in most jurisdictions under a “lustful disposition” rule220United States v. Castillo, 140 F.3d 874, 881 (10th Cir. 1998) (citing Thomas J. Reed, Reading Gaol Revisited: Admission of Uncharged Misconduct Evidence in Sex Offender Cases, 21 Am. J. Crim. L. 127 (1993)). (codified in the federal rules at Federal Rule of Evidence 415).221Fed. R. Evid. 415. Prior sex crimes evidence is admissible on the inference that the defendant has a propensity to commit sexual assault—which is different from a propensity to have consensual sex with a member of the same gender. This significantly raises the probative value to the point where it likely outweighs the risk of prejudice created by the fact that the prior victims were of the same gender.

C. A Long Look in the Mirror

Any evidence that survives both a logical relevance and a character analysis is then subject to a legal relevance balancing test. Under the legal relevance rule, codified in the federal rules at Federal Rule of Evidence 403, evidence may be excluded when the probative value is substantially outweighed by the risk of unfair prejudice, inter alia.222Fed. R. Evid. 403. Although exclusion is discretionary, public policy strongly favors the exercise of that discretion to exclude evidence of defendant sexual orientation in order to combat inherent heterosexism. This proposal requires that judges be mindful not only of the many factors that undermine the probative value of defendant sexual orientation in a sexual assault case, but also of the continued prevalence of anti-gay and anti-bisexual stereotypes, sexual prejudice, and heterosexism.

Judges must carefully weigh the probative value, paying particular attention to the complexities of sexual orientation, against the risk of unfair prejudice. That risk includes not only the potential that a fact-finder will make the forbidden inference that gay, lesbian, and bisexual people are more likely to commit sexual violence, but also the more mundane risk that the fact-finder will simply misuse the evidence by overestimating its probative value based on a misunderstanding of the phenomenon of sexual orientation. A judge who holds anti-gay stereotypes may be less likely to exclude the evidence precisely because she or he will tend to overestimate the probative value and underestimate the potential for unfair prejudice. Becoming familiar with the research discussed in Section II.B above is an important first step towards addressing this problem; and scrupulous adherence to the evidentiary analysis above should nevertheless result in most of this type of evidence being excluded at the character stage.

However, this problem did not arise exclusively from bad decisions by biased judges. Perhaps the biggest challenge for many judges, particularly those who view themselves as supportive of sexual equality, lies in accurately valuing the risk of unfair prejudice.223See supra Section III.A. Implicit bias prevents judges from recognizing their own heterosexism, and strong forces make it difficult to admit the bias that exists in one’s own community.224Cf. supra Section III.A. In order to solve the problem of improper admission and misuse of defendant sexual orientation evidence in sexual assault cases, judges must commit themselves to fully accepting that heterosexism and anti-LGBTQ forces still pervade American society and compromise the fairness and integrity of our justice system.

Conclusion

As the United States lurches unsteadily towards equality for gay, lesbian, and bisexual people, we are still plagued by sexual bias and heterosexism, which negatively impact the justice system and society. This Article has focused on dismantling the double-standard created by courts’ handling of evidence of defendants’ sexual orientation in sexual assault cases as one means of mitigating sexual bias and heterosexism in the courts. As scientific and theoretical understanding of sexual orientation and sexual assault continue to evolve, it is increasingly clear that the link between sexual orientation evidence and sexual assault perpetration is, at most, highly attenuated. The probative value of such evidence is decidedly low because sexual orientation is multifaceted, dynamic, and exists on a continuum rather than in discrete categories, and because sexual assault is heterogenous, complex, and multi-factorial. Moreover, evidence of a defendant’s sexual orientation is replete with risk of unfair prejudice based on sexual prejudice and negative stereotypes about gay, lesbian, and bisexual people.

In striving for fair justice system outcomes for gay, lesbian, and bisexual people, courts must be highly skeptical of this evidence and use a disciplined approach in applying existing rules of evidence, including logical relevance, character rules, and the legal relevance balancing test. Because most evidence of defendant sexual orientation in sexual assault cases is relevant only through a propensity theory, that evidence must be subject to analysis under the character evidence rules. Special attention must be paid to surfacing propensity theories hidden within otherwise permissible theories such as motive or intent. When properly applied, the existing rules of evidence will result in exclusion of defendant sexual orientation in nearly all cases. This approach calls upon judges to actively work to overcome both explicit and implicit biases in order to fully account for the pernicious anti-gay and anti-bisexual forces that pervade American society and color fact-finders’ abilities to accurately and fairly weigh sexual orientation evidence.