Animals in Suits: What 121 Primate Species Reveal About Why Sexual Coercion Keeps Working in Congress and Hollywood

By Courtney Turrin · September 1, 2026 · 16 min read read

Male primates in suits sitting around a conference table — AI-generated image showing male primates in suits sitting around a conference table

New research on primate species suggests sexual coercion is not simply a matter of animal instinct, challenging common assumptions.

This piece discusses sexual violence, including prevalence data and references to specific allegations. If you need support, RAINN's National Sexual Assault Hotline is available 24/7 at 1-800-656-4673.

CNN and NPR reported in August 2026 that the U.S. House Committee on Ethics had opened an investigation into Representative Jimmy Gomez over allegations of sexual misconduct. Gomez has said his conduct was consensual. Attorneys for one woman told the committee otherwise.

When stories like this hit the headlines, the reflex is to reach for biology. Men are animals. Something base and ancient drives this behavior. The business suit was only ever a costume over the ape.

The comparative evidence doesn’t support this argument.

Nearly Half of Women Report Contact Sexual Violence. The Systems Built to Record It Capture Almost None.

The Centers for Disease Control and Prevention's National Intimate Partner and Sexual Violence Survey finds that nearly half of women and more than one in six men in the United States have experienced some form of contact sexual violence during their lifetimes. More than one in five women and one in 31 men have experienced completed or attempted rape.

Despite the prevalence of sexual violence, the vast majority goes unreported. The Equal Employment Opportunity Commission's published research finds that roughly 90% of people who report experiencing workplace harassment never take any formal action. No lawsuit. No internal complaint. Nothing that produces a record.

Whenever the topic of sexual violence and reporting is raised, the incidence of false reporting always follows, despite the evidence showing its rarity. Psychologist David Lisak and colleagues examined all 136 sexual assault cases reported to a major Northeastern university across a decade and coded 5.9% as false. Reviewing seven other methodologically rigorous studies alongside their own, they placed the prevalence of false allegations between 2% and 10%. End Violence Against Women International, applying stricter criteria requiring either a credible admission or strong evidential grounds, narrows the range to 2% to 8%.

That range may still overestimate false reporting. Criminologist Cassia Spohn and colleagues studied how the Los Angeles Police Department decided which sexual assault reports to close as "unfounded," the category much of the false-report literature draws on, and found that the strongest predictor of an “unfounded” decision was the complainant recanting, not evidence that the assault didn't happen. A case can be marked unfounded because a victim recants under pressure, because a detective doubts her, or because the investigation stalls, none of which proves the report was false. Folding those cases into the false-report count inflates the rate (Figure 1).

[datawidget key="prevalence-reporting-false"]

Lifetime prevalence surveys, workplace complaint behavior, and the disposition of police reports are measuring different factors, but they share a direction. Assaults that should count as real get lost to non-disclosure, to institutional inaction, and to police paperwork that treats an unprosecutable case as a false report. The system built to record this behavior undercounts it at nearly every stage, and the undercount is structural rather than accidental. That's a fact about how human institutions handle disclosure. It isn't an explanation for why the coercion happens in the first place, and the reflexive explanation for that, that men are simply acting on animal instinct, is testable. The nonhuman primate record is where to test it.

Males Clearly Dominate Females in 17% of Primate Populations. Females Dominate in 13%.

In July 2025, behavioral ecologist Élise Huchard and colleagues published the largest quantitative survey of intersexual power in primates ever assembled. Working across five years, the team compiled aggression and dominance data from 253 populations representing 121 primate species, spanning monkeys, lemurs, tarsiers, and lorises.

The headline finding overturns a premise that shaped the field for decades. Societies in which males win nearly all aggressive encounters against females are rare. Males won more than 90% of male-female contests in fewer than one in five of the populations studied. Clear male dominance appeared in 17% of populations, clear female dominance in 13%, and the remaining 70% showed shared or ambiguous power (Figure 2).

[datawidget key="dominance-distribution"]

The team then tested what predicts the distribution. Female-biased dominance emerges where females have substantial control over reproduction: in monogamous species, in species where males and females are the same size, and in arboreal species. Male-biased dominance prevails where species are terrestrial, where males are larger or better armed, and where males mate with multiple females.

Huchard's explanation of the mechanism is precise. Primate males gain power through physical force and coercion. Female power runs through different channels, primarily reproductive control over matings.

The study's authors go further. Humans do not carry the full trait set that characterizes species where males strictly dominate females. The human profile sits closer to species with negotiated relationships, where individuals of either sex can become dominant. On the evidence, arguments presenting human patriarchy as a primate inheritance are misdirected.

So the "men are apes" defense fails on its own terms. It is not that humans are behaving like other primates. It is that humans are behaving unlike most of them.

Coercion Rises and Falls With How Much It Costs the Male Who Attempts It

Across the species with the best field data, coercion tracks how costly it is for a male to attempt it: whether he can physically overpower a female in the first place, and whether she has other animals nearby who will make him pay for trying (Figure 3).

[datawidget key="species-coercion-table"]

Orangutans sit at one extreme. Both flanged and unflanged males force copulations, producing what anthropologist Cheryl Knott has described as among the highest rates of forced copulation documented in any animal. Orangutans are also semi-solitary. Females are alone by default, and the ecological cost of association keeps them that way. Their solitary habit also means females have no allies nearby to call on.

Chimpanzee females disperse at maturity into communities where they have no kin and few established allies. Sexual coercion in chimps is routine, and primatologist David Watts's 2022 study found that male dominance rank and rates of aggression toward fertile females each predicted higher mating success, measured by copulation rates and by the share of copulations a male achieved.

Hamadryas baboons make the mechanism explicit. Primatologists Larissa Swedell and Amy Schreier documented conditioning, coercion, and control as an integrated male strategy: females are coercively transferred between social units by males, which restricts female mate choice and bonding among female kin. Swedell and colleagues later quantified the cost. Of coercive takeovers involving females with young infants, 67% were followed by infant mortality.

Chacma baboons show the same mechanism operating across a longer timescale. Alice Baniel and colleagues tracked two wild groups in Namibia across four field periods. Males directed unprovoked aggression at fertile females, making male violence the primary source of injury to cycling females. The aggression did not increase mating success immediately. It increased the probability that the same male would guard the same female during her peak receptivity weeks later. Intimidation with a lag.

Female alliance is not a universal counterstrategy to male coercion in nonhuman primates. Primatologist Nikolaos Smit and colleagues found that in mandrills, socially integrated females faced more sexual coercion than isolated ones, which runs counter to a simple story where female bonds are protective.

At the other extreme are bonobos. Behavioral ecologist Martin Surbeck and colleagues analyzed 30 years of observation across six wild communities. Of 1,786 recorded conflicts between the sexes, females won 61.5%. Of coalitionary aggression, 85% was directed at males. Male bonobos are larger and stronger. Coalitions form within seconds, typically including three to five individuals. When the match up becomes one on three or one on five, the size differential between males and females stops mattering (Figure 4).

[datawidget key="bonobo-conflict-outcomes"]

Anthropologist Karen Strier has studied the same Brazilian population of muriquis for more than four decades. Muriquis show no dominance hierarchy and almost no agonistic behavior. Males and females are the same body size with the same canine size, so no male has the physical means to reliably overpower a female. They mate in full view of one another without threats or interference. Strier's genetic work quantified what the behavioral observation suggested: across 22 infants born over three years, paternity was distributed across 13 fathers, and the most successful male sired 18% of offspring, a fraction of what top males achieve in high-coercion species like chimpanzees.

Muriquis are also patrilocal: males stay in their natal group for life, and females leave before puberty to join a different one, the same pattern found in chimpanzees, bonobos, and roughly 70% of documented human societies, according to anthropology's Ethnographic Atlas, where wives have historically moved to live among their husbands' kin.

Compare muriquis and bonobos to chimpanzees, the clearest case of both physical size and coalition access lining up in males' favor. Bonobo males are just as large as chimpanzee males, but bonobo females have fast coalition backup whereas chimpanzee females don't, and coercion is much lower in bonobos. Muriqui females are just as isolated as chimpanzee females, but muriqui males have no size advantage to exploit and chimpanzee males do, and coercion is again much lower in muriquis.

This suggests that if we take away either condition, a capable male or an uncontested one, sexual coercion declines. Mandrills are the exception this framework hasn't resolved: socially integrated females there face more coercion, despite real backup. Strier's own conclusion from the work was that there is no such thing as a typical primate.

No Nonhuman Primate Has a Nondisclosure Agreement

Human institutions did not merely fail to reduce coalition latency. They built instruments to extend it.

The clearest example is the one that started a modern reckoning. Reporting by Jodi Kantor and Megan Twohey in October 2017 established that Harvey Weinstein, over nearly three decades, reached at least eight settlements with women who had raised allegations against him. The recipients included an assistant in New York in 1990, an actress in 1997, an assistant in London in 1998, and an Italian model in 2015. Weinstein has since been prosecuted in New York and California.

Eight women, eight separate nondisclosure agreements (NDAs), each one legally barred from speaking out and therefore from finding out about the other seven. That is not an incidental byproduct of settlement. It is the product being purchased.

Congress operates its own version with public money. The Office of Congressional Workplace Rights approved 349 awards and settlements between 1996 and December 2018. CNN reported that more than $550,000 in confidential sexual harassment settlements involving six former members surfaced only after the office was subpoenaed. Records predating 2004 were destroyed, suggesting that the taxpayer dollars spent on awards and settlements for sexual harassment cases involving members of Congress exceed what's been substantiated (Figure 5).

[datawidget key="settlement-machinery"]

The Epstein files show the same architecture at a scale that required an act of Congress to breach. The Epstein Files Transparency Act was signed on November 19, 2025. The Justice Department has since published roughly 3.5 million pages, more than 2,000 videos, and 180,000 images. Tens of thousands of pages were subsequently removed from the public repository. A federal judge has pressed the department to justify its redactions and warned about contempt proceedings, but at the time of publication, many of the files remain undisclosed (Figure 6).

[datawidget key="epstein-disclosure"]

None of this material was made public through the ordinary operation of the institutions holding it. It was released under statutory compulsion, and it is still contested.

A bonobo coalition costs nothing to form. A human coalition, under this architecture, requires each participant to breach a contract she signed while at her most vulnerable, without knowing whether any other survivors exist and would breach theirs.

MeToo Raised Reporting 10% Before Any Law Changed

The question of whether the MeToo movement, which began in 2006 and went viral in 2017, accomplished anything has an empirical answer.

Economists Ro'ee Levy and Martin Mattsson examined reporting of sex crimes across 30 countries in the Organisation for Economic Co-operation and Development (OECD), a group of mostly high-income democracies whose crime-reporting data is standardized enough to compare across borders. In the six months after October 2017, reporting rose approximately 10%. The effect persisted for at least 15 months, and arrests increased over the longer run. The increase in reporting preceded the legal changes that followed it, which means the movement worked through social norms rather than statute (Figure 7).

[datawidget key="metoo-reporting-effect"]

Read against the primate frame, this is a coalition-latency result. MeToo lowered the cost of being the first person to speak by demonstrating, in real time, that others would follow.

In a separate finding worth flagging, sociologist Michael Rosenfeld's October 2024 analysis of U.S. Merit Systems Protection Board survey data found that sexual harassment reported by women in the federal workforce fell by more than 50% between 1987 and 2016. Rosenfield attributes this decline partly to workforce demographics and generational turnover, and it predates the reporting increase Levy and Mattsson measured by more than a year. Rosenfeld's data establishes that reported harassment rates move for structural reasons that have nothing to do with any one movement, which is useful context before crediting MeToo alone for the shift in reporting behavior after 2017.

Encrypted Matching Lets a Coalition Form Before Anyone Has to Go Public, but Only Catches Repeat Offenders.

In 2018, law scholar Ian Ayres proposed a targeted remedy for repeat-offender NDAs built on a mechanism called an information escrow. The design allows a person to deposit an accusation that stays sealed unless and until a matching accusation arrives, at which point both are released to a designated counselor.

The nonprofit Callisto built it for university populations specifically. Callisto Vault lets a survivor enter an encrypted identifier for a perpetrator. If a second user enters the same identifier, regardless of institution, a match triggers and a confidential legal options counselor contacts each person separately. Callisto describes the system as a secure alternative to a whisper network. Access opened to anyone with a .edu email address at any U.S. college in fall 2023 (Figure 8).

[datawidget key="information-escrow-process"]

Structurally, it is a bonobo coalition rendered in cryptography. Nobody goes first alone, because nobody goes first at all.

Callisto's reported outcomes show that approximately 15% of entries match with at least one other survivor, and more than half of matched survivors choose to coordinate action. Survivors in 49 states have created accounts, producing cluster matches involving three or more people and matches spanning opposite coasts.

However, matching escrows assume repeat offending. Callisto's public case cites figures like 90% of campus assaults being committed by serial perpetrators, tracing to a 2002 study by David Lisak and Paul Miller. In 2015, psychologist Kevin Swartout and colleagues applied latent class growth analysis to the two largest longitudinal datasets on campus sexual violence and found trajectory patterns inconsistent with a serial rapist model. Of the men who reported perpetrating, roughly a quarter followed a trajectory matching serial perpetration. The paper explicitly noted that the scientific foundation for the serial pattern assumption rests largely on the single Lisak and Miller source.

The Swartout paper was corrected twice in 2015 for missing-data inconsistencies affecting two cases. Georgia State University's research integrity office found no misconduct, and the authors maintain that their conclusions stand. Callisto's administrator materials concede that studies of serial perpetration are limited (Figure 9).

[datawidget key="escrow-premise-tension"]

If most perpetration is not serial, matching detects a smaller share of it than the advocacy figures suggest. The mechanism still lowers coalition latency for the cases it catches, but it is a narrower instrument than is needed.

Virginia Banned the NDA That Rarely Silences Anyone

Virginia Code § 40.1-28.01, enacted in 2019, voided nondisclosure and confidentiality provisions concealing details of sexual assault claims. House Bill 1895, signed March 26, 2023, and effective that July, extended the ban to sexual harassment claims and to non-disparagement provisions. Any such provision is against public policy and unenforceable.

The statute reaches agreements an employer requires of an employee or prospective employee as a condition of employment.

The phrase "as a condition of employment" introduces an important point of contention. Attorneys at Jackson Lewis read the law as not restricting agreements with former employees. Writing in Virginia Lawyer, practitioners concluded it does not appear to apply to post-employment severance or separation agreements, which makes it narrower than comparable laws in other states. Attorneys at Troutman Pepper Locke read HB 1895 more expansively, as arguably covering any agreement entered during employment including settlement of claims, and note the construction remains untested (Figure 10).

[datawidget key="virginia-nda-coverage"]

Settlement nondisclosure agreements are, by nature, executed at exit. Under the narrower reading, Virginia has prohibited the version of the instrument that rarely silences anyone and left intact the version that reliably does. Until a court resolves the scope, Virginia employers and employees are operating on a statute whose central term has two defensible meanings.

Two Conditions Explain Primate Coercion. Humans Only Get to Change One.

The nonhuman primate record does not settle an argument about human nature. What it shows, across species that vary in size, mating system, and social structure, is that sexual coercion is not fixed by anatomy alone. It shows up reliably where a male has the physical means to force the issue and the targeted female has no fast way to make him pay for it. Take either condition away, a heavily dimorphic species with rapid backup, or a species with no size advantage to begin with, and coercion drops even though nothing else about the species' biology has changed.

Humans are not a special case here. Like most primates, humans carry real, if moderate, sexual size dimorphism: the physical-capacity condition is generally present in humans the way it is in chimpanzees, not absent the way it is in muriquis. What varies, across cultures, workplaces, and legal systems, is the other condition: whether the person targeted has a fast, credible way to get backup. That is the condition every primate species in this comparison leaves to social structure. It is also the condition human institutions have taken a direct hand in shaping.

No nonhuman primate has ever drafted a nondisclosure agreement. No chimpanzee community has destroyed its own records of past offenses. No bonobo coalition needs a subpoena to form. Bonobos show that fast backup is enough to neutralize a real size advantage, the same size advantage humans carry. Human institutions work in the opposite direction. The settlements, the sealed committee records, the files that took an act of Congress to unlock, were built to make sure that backup can't be reached in time, can't be believed, or can't act before the claim expires. That isn't evolution. It's engineering, and of every species in this comparison, humans are the only one that built it on purpose.