Legal scholarship has documented a specific mechanism for how women’s accounts get discounted in court, and a specific set of interventions that measurably interrupts it.
Legal scholarship has a name for what looks, from the outside, like a woman being “too emotional” or “not credible enough” to be believed: the credibility discount. It is a documented, measurable pattern in which women’s accounts are discounted by default in legal proceedings, regardless of the underlying facts, while men accused of wrongdoing benefit from a corresponding credibility boost. The research also identifies what closes the gap. Advocates, counsel, and corroborating witnesses present in the room change legal outcomes in ways that show up in the data. That is a procedural fact, not an indictment of the women who go through hearings without one.
The Mechanism, Named
The pathologizing of women’s emotional distress in legal and medical settings is not a new phenomenon dressed up in modern language. Deborah Tuerkheimer, a former sex crimes prosecutor now on the Northwestern law faculty, has traced how a woman’s account of harm gets filtered through what she calls a credibility complex before anyone reaches the underlying facts. Her framework, first laid out in the University of Pennsylvania Law Review under the title Incredible Women: Sexual Violence and the Credibility Discount, describes a structural bias that discounts the credibility of the people who report harm, most often women, while inflating the credibility of the people accused of causing it.
The mechanism operates before fact finding, not after it. Police decide whether a report becomes a case. Intake staff decide whether a petition gets a hearing date. Prosecutors decide whether to charge. Judges decide whether an account is credible enough to act on. Tuerkheimer’s later scholarship on the “new rape paradigm” documents how each of these decision points can apply the discount independently, so that by the time a case reaches a courtroom it has already survived, or failed to survive, several rounds of gatekeeping that had nothing to do with a jury or judge weighing evidence.
What makes this useful to document rather than simply infuriating to describe is that it is testable. The discount shows up in controlled experiments, not just in anecdote, which means it can be measured, targeted, and in principle, corrected.
The Reversal Tactic: DARVO as Procedural Weapon
Psychologist Jennifer Freyd coined the term DARVO, deny, attack, reverse victim and offender, to describe how an accused person can deflect accountability by denying the conduct, attacking the accuser’s credibility, and then reframing themselves as the true victim of a false accusation. It is not a courtroom tactic exclusively. It shows up in HR meetings, family disputes, and public statements. But its legal-proceeding version is the most consequential, because it is aimed directly at the fact finder.
Two controlled experiments led by Sarah Harsey and Freyd measured what DARVO actually does to an observer’s judgment. In the first, 316 university students read vignettes describing a sexual violence scenario, with or without DARVO tactics present. Participants exposed to DARVO rated the accused as less abusive and less responsible, and rated the true victim as less believable and more responsible for what happened, even though the underlying facts in the vignette had not changed. A follow up study replicated the pattern with a comparison group of 230 undergraduates and found the same directional effect.
Nothing about the facts changed between conditions in these studies. Only the accused party’s rhetorical posture changed. That is the entire finding: denial plus attack plus reversal moves the fact finder’s judgment, independent of what actually happened.
Freyd’s later work extends the concept to institutions themselves. When an institution, a police department, a court, an employer, adopts the same deny-attack-reverse posture toward the person raising a claim rather than toward the underlying conduct, she calls it institutional DARVO. It is the same mechanism, scaled up from one accused individual to an entire gatekeeping apparatus.
The No-Win Emotional Calibration
If DARVO is something done to a woman’s account by an opposing party, the emotional victim effect is something the system does on its own. A 2019 meta-analysis by Faye Nitschke, Blake McKimmie, and Eric Vanman, published in Psychological Bulletin, synthesized the existing research on how a complainant’s visible emotion affects perceived credibility in sexual assault cases. The finding is not simply that emotional women are disbelieved. It is that there is no safe emotional register. Complainants who fail to display enough visible distress are read as less credible, on the theory that a real victim would be more upset. Complainants who display significant visible distress risk being read as unstable, manipulative, or performing.
Tuerkheimer’s more recent scholarship situates this inside a longer history. Long before “not credible” became the operative courtroom language, the same function was served by explicitly medical vocabulary, hysteria diagnoses rooted in centuries-old theories about women’s bodies and minds. The vocabulary has modernized. The function, discrediting a woman’s account by discrediting her composure, has not.
For a woman inside an actual hearing, this is not an abstract finding. It means there is no correct way to testify. Too calm reads as fabricated. Too upset reads as unreliable. The bind is not a personal failure of self-presentation. It is a documented structural trap.
When a woman first discloses harm, to police, an intake worker, a hotline volunteer, or an employer, the emotional victim effect is already active before any fact finding happens. Observers judge her account partly on demeanor: too little visible distress reads as fabrication, too much reads as instability.
Before a case reaches a judge, it passes through gatekeepers: police deciding whether to write a report, prosecutors deciding whether to charge, court staff deciding whether a petition gets a hearing date. The credibility discount compounds at every handoff.
This is where DARVO does its most measurable work. When an accused party denies, attacks, and reverses victim and offender roles inside a hearing, fact finders rate the true victim as less believable and more responsible for what happened.
A granted order or a favorable ruling is not the end of the discount. Research on civil protection orders finds that whether a violation gets taken seriously, and whether a longer order gets renewed, again depends partly on courtroom actors and documentation quality.
Clutch Justice’s course library teaches judicial accountability, FOIA practice, and court literacy from the ground up, the same methodology behind every investigation on this site.
Browse the Courses ?What Actually Closes the Gap
The research does not stop at documenting the problem. It also identifies what interrupts it. A 2019 study in the Journal of Family Violence by Alexa Bejinariu, Emily Troshynski, and Terance Miethe examined 303 protection order hearings in a large Southwestern jurisdiction, coding the presence of formal actors, a hearing officer, legal counsel, a courtroom translator, and informal actors, a victim advocate, family members, friends. The finding: whether a petitioner obtained an order, and for how long, depended in part on who else was present in the courtroom with her. The authors’ policy recommendation follows directly from the data: states should fund non-profit legal aid and victim advocacy programs, not as a courtesy service, but as an intervention with a measurable effect on outcomes.
A separate line of research on juror knowledge reinforces the same structural point. Studies of child sexual assault trials have found that when jurors have more accurate knowledge about how disclosure and delay actually work, their perceived credibility of both the complainant and a corroborating witness rises, and conviction rates move accordingly. The lever is not the complainant becoming more persuasive. It is the fact finder’s isolation being broken, either by another person’s account or by better information about what a credible disclosure actually looks like.
Put together, the pattern is not that women need a man in the room to be believed, as it is sometimes crudely put. It is that the credibility discount is calibrated to operate on an isolated claimant, one word against another, and the documented interventions that interrupt it, an advocate, counsel, a corroborating witness, an informed fact finder, all work by breaking that isolation. The gender of the person breaking it is incidental to the mechanism. The isolation is the target.
Grades reflect documented structural exposure to the credibility discount identified in the research above. They are not an evaluation of any individual court, judge, or case.
Where the “Just Bring a Witness” Fix Breaks Down
The honest counterargument to everything above is this: if presence of an advocate or witness is what changes outcomes, doesn’t that just mean the system rewards whoever has more resources and connections, rather than whoever is telling the truth? And isn’t telling women to bring a witness itself another burden placed on the person who was harmed, while the underlying disbelief default goes unexamined?
That critique is well founded, and it shows up directly in the scholarship. A 2023 study in the Journal of Interpersonal Violence by Ellen Reeves, Kate Fitz-Gibbon, Silke Meyer, and Sandra Walklate, titled Incredible Women: Legal Systems Abuse, Coercive Control, and the Credibility of Victim-Survivors, documents how legal systems themselves become tools that abusers weaponize against victims, filing motions, requesting hearings, and using the litigation process itself to extend control long after a relationship ends. In that context, a presence-based fix that is not funded or guaranteed does not close the credibility gap. It relocates it onto whoever cannot afford or access an advocate, which frequently tracks the same lines of income, geography, and language access that already predict disadvantage everywhere else in the system.
This is the distinction that matters for anyone trying to act on this research rather than just find it validating. The documented interventions are not a prescription for what individual women should do differently, show up with a friend, hire a lawyer, find a corroborating witness. They are a diagnosis of where the system’s gatekeeping function breaks down. That diagnosis points toward structural, funded fixes. It does not point toward better individual performance under an unfair test.
What Fixing It Would Actually Require
The research base above is not Michigan-specific. It is drawn from multi-jurisdictional academic literature, which means the mechanism it documents is not confined to one state’s courts. Whether Michigan’s own protection order and family court data show the same presence effect is an open, testable question, and one that a jurisdiction interested in its own accountability could answer directly by tracking the same variables these researchers did.
What the existing research already supports is a specific set of structural changes, not a general call for more sensitivity or more awareness.
None of this requires believing women by default. It requires building a system that no longer defaults to disbelieving them, and the research on how to do that already exists. What is missing is the funding and the statute, not the evidence.
Continue Your Investigation
If this reporting raised more questions, use the Clutch Justice ecosystem to keep going.
