The Epstein operation was extraordinary in its documentation, not in its structure. The recruitment pyramid, the institutional cover, the silenced victims, the delayed accountability — that mechanism is documented in five U.S. institutional sexual abuse cases that preceded it, and in the population-level prevalence data on child sexual abuse in this country and globally. This is the data, with sources, in five sections.
1. U.S. prevalence
The Centers for Disease Control:
"At least one in four girls...in the United States experience child sexual abuse."
"One in 20 boys in the United States experience child sexual abuse."
"About 90% of child sexual abuse is perpetrated by someone known and trusted by the child."
CDC, "About Child Sexual Abuse," cdc.gov
The DOJ Bureau of Justice Statistics' foundational 2000 study, still the most-cited perpetrator-relationship breakdown, parsed that further: of juvenile cases reported to law enforcement, ninety-three percent of victims knew the perpetrator — fifty-nine percent acquaintances, thirty-four percent family members, seven percent strangers. (RAINN's summary of DOJ BJS, "Sexual Assault of Young Children as Reported to Law Enforcement," 2000.)
The adult-women lifetime baseline is consistent. CDC's National Intimate Partner and Sexual Violence Survey (2016/2017 cycle, the most recent comparable methodology): fifty-four-point-three percent of women in the U.S. have experienced contact sexual violence in their lifetime; twenty-six-point-eight percent — one in four — have experienced completed or attempted rape; seventy-nine percent of female rape survivors were first raped before age twenty-five; forty percent before age eighteen. Three-quarters of sexual assaults go unreported. Less than one-third of rape victims report to police.
Disclosure delay is the structural finding under all of it. CHILD USA's 2024 synthesis: the average age of disclosure is fifty-two; over seventy percent of CSA victims do not disclose within five years; only twenty-one-point-three percent disclosed within one month; one in five survivors never disclosed at all; up to a third of child victims will die without telling anyone. (CHILD USA 2024 factsheet.)
Read those numbers as a clinical operating reality, not as headline statistics. The country is not seeing the cases as they happen. The country is seeing the residue — the survivors who reach disclosure threshold decades later. The visible pile is the iceberg's tip.
If I see twenty adolescent girls in a clinic day, five of them — statistically, by the CDC baseline — have been sexually abused as children. Maybe one will ever say so out loud to a clinician. The average age of disclosure is fifty-two. The patients I see in their twenties and thirties have not, on average, reached the disclosure threshold yet. The harm is not invisible. It is below the surface of the encounter. The clinical job is to read the residue: the patient who flinches at the pelvic exam, who refuses the speculum, who picks the contraceptive nobody can see, who decompensates beyond reason at the foot of a delivery bed. The one-in-four number is not a statistic when it sits in front of you. It is the room.
2. Trafficking-specific data
The National Center for Missing & Exploited Children, 2024:
"In 2024, the CyberTipline received 20.5 million reports, including nearly 63 million files."
"Reports involving generative AI surged by 1,325%, climbing from 4,700 in 2023 to 67,000 in 2024."
"The CyberTipline received more than 546,000 reports concerning online enticement — a 194% increase compared to 186,000 reports received in 2023."
"Child sex trafficking reports were up 55% from 2023."
NCMEC, "NCMEC Releases New Data: 2024 in Numbers." missingkids.org/blog/2025
NCMEC's 2025 first-look numbers, released early 2026: twenty-one-point-three million reports; one-point-four million reports of online enticement (a one-hundred-fifty-six percent increase year over year); trafficking reports up three-hundred-twenty-three percent year over year; more than fifty-three thousand reports escalated as urgent or involving a child in imminent danger. (NCMEC 2025 first-look.)
The Polaris Project ran the National Human Trafficking Hotline 2007–2025. Polaris's 2024 final-year report: eleven thousand, nine-hundred ninety-nine trafficking situations identified; twenty-one thousand, eight-hundred sixty-five victims; thirty-two thousand, three hundred nine total signals to the hotline, including just over eight thousand contacts from victims themselves. (Polaris Project.)
Both data sources are reporting the visible portion. CyberTipline reports decreased slightly between 2023 and 2024, attributable primarily to Meta's expansion of end-to-end encryption, which suppressed reporting volume from Meta platforms. Fewer reports do not mean fewer crimes. They mean less detectability.
From the Architecture of Harm Chapter 9: healthcare providers identify fewer than one percent of trafficking victims, despite the fact that eighty-eight percent of trafficking survivors have contact with a healthcare facility during their exploitation — an average of four-point-five clinical encounters before anyone asks the right question. The trafficking apparatus is moving through the healthcare system at high volume. The healthcare system is not seeing it.
3. Institutional pattern — one mechanism, five costumes
Across five independent institutions, the mechanism is identical: adults with proximity and institutional authority used the institution's trust architecture (parish, locker room, scout troop, coaching role, ministry) to access children; the institution kept internal records of credible allegations; the institution did not act on those records; disclosure was delayed by decades; settlement followed only after independent journalism, court order, or government investigation forced the records into public view. Aggregate institutional liability across the five cases below exceeds ten billion dollars.
Catholic Church
From the U.S. Conference of Catholic Bishops' annual audits (aggregated):
"As of April 2026, the Catholic Church has paid over $5 Billion in clergy sexual abuse settlements and legal fees in the U.S. Additionally, nearly 11,000 allegations of sexual abuse have been brought against over 4,000 priests."
Sokolove Law tracker, citing USCCB audit data. sokolovelaw.com
The Boston Globe Spotlight investigation broke in January 2002. By year-end, six hundred stories had been published. (Pulitzer archive.) The 2018 Pennsylvania Grand Jury Report identified more than three hundred priests credibly accused of abuse, with over one thousand identified victims, across seventy years of records. The 2017 Australian Royal Commission found the same pattern in another country, in the same Church. Same mechanism. Same record-keeping practice. Same institutional response.
USA Gymnastics / Larry Nassar
Confirmed victims: more than five hundred. (Wikipedia summary; USAG investigation records.) The Indianapolis Star broke the story in September 2016 after an eight-month investigation. Cumulative institutional liability — Michigan State University, USA Gymnastics + U.S. Olympic & Paralympic Committee, and the U.S. Department of Justice (for the FBI's investigative failure):
"This settlement will resolve the last remaining legal claims against the state and federal institutions who failed to protect more than 500 girls and women from Nassar and bring their total liability to almost $1 billion."
U.S. Department of Justice press release, April 2024. justice.gov
Mechanism artifact: the FBI slow-walked its investigation for fourteen months after USAG's 2015 report. Nassar continued abusing during that window. The Karolyi Ranch was the locked-room access point. "Medical treatment" was the cover.
Boy Scouts of America
"$2.46 Billion settlement fund for more than 82,000 survivors of sexual abuse, making it one of the largest child sexual abuse settlements in U.S. history."
"Leaders maintained internal records known as the 'perversion files' for over 100 years to keep track of sexual abusers instead of reporting them to the police."
Sokolove Law BSA tracker. sokolovelaw.com
The Supreme Court declined to hear a challenge to the settlement on January 12, 2026. One-point-six-five billion in escrow released February 11, 2026; over thirty-nine thousand survivors paid more than three-hundred sixteen million dollars by April 2026. Mechanism artifact: the BSA "perversion files." A century-long internal archive of credible allegations, maintained by the institution, used to manage personnel — not to report crimes. The structural analog to Epstein's "little black book."
Penn State / Jerry Sandusky
Convicted June 22, 2012 on forty-five counts of child sexual abuse; sentenced to a minimum of thirty years. Eight young men testified that Sandusky had befriended them through his charity, the Second Mile, and abused them in his home, in the Penn State football locker room, and on road trips with the team. NCAA sanctions: four-year bowl ban, $60 million fine, 112 vacated wins. The Freeh Report (July 12, 2012) found that head coach Joe Paterno, President Graham Spanier, Athletic Director Tim Curley, and Senior Vice President Gary Schultz were complicit in "conceal[ing] Sandusky's activities from the Board of Trustees, the University community and authorities" and showed "total and consistent disregard" for child sex abuse victims. (Washington Post, July 12, 2012.) Mechanism artifact: the Second Mile charity. An institutionally blessed access pipeline. The structural parallel to Epstein's "modeling" recruitment channel.
Southern Baptist Convention
The 2022 Guidepost Solutions investigation, commissioned by the SBC Executive Committee, produced a 209-page report released May 22, 2022:
"The report detailed how the Southern Baptist Convention's Executive Committee mishandled allegations of sex abuse, stonewalled numerous survivors and prioritized protecting the SBC from liability. Additionally, the report also stated that known abusers were allowed to keep their positions without informing their current church or congregation."
NPR summary of the Guidepost report. npr.org
The Executive Committee maintained a secret database of credibly accused abusers — more than seven hundred entries from cases spanning 2000 to 2019. Released by Guidepost. The U.S. Department of Justice opened an investigation in August 2022. Mechanism artifact: the secret database. The same institutional archive practice as the BSA perversion files and the Catholic Church diocesan personnel files.
The list is long because the mechanism is ordinary. Five independent institutions, five different American sectors, five secret archives of credibly accused abusers, five institutional protection patterns, five exposures forced only after the historical record became too heavy to suppress. The Epstein operation is the sixth — the institution of Epstein himself, with the same access architecture, the same internal record-keeping, the same external protection function performed by a network of powerful men, the same eventual exposure forced by independent journalism and survivor testimony. What is different about the Epstein case is what got documented. The flight logs. The email cache. The black book. The mechanism is identical. The paper trail is what is unusual.
4. Global context
The 2025 Lancet meta-analysis of global sexual violence against children is the most authoritative current global estimate:
"The global age-standardised prevalence of sexual violence against children was 18.9% for females and 14.8% for males in 2023."
"The first experience of sexual violence among adolescents and young people occurred before the age of 18 years for 67.3% of female and 71.9% of male survivors."
Lancet, "Prevalence of sexual violence against children and age at first exposure: a global analysis by location, age, and sex (1990–2023)," 2025. thelancet.com
UNICEF's 2024 update: six hundred fifty million girls and women alive today were subjected to sexual violence as children — one in five globally — including over three hundred seventy million (one in eight) who experienced rape or sexual assault. Between four hundred ten and five hundred thirty million boys and men (around one in seven) experienced sexual violence in childhood, including two hundred forty to three hundred ten million (around one in eleven) who were raped or sexually assaulted. (UNICEF press release 2024.)
The comparison. The global rate for girls is roughly one in five. The U.S. rate is one in four. The United States is worse than the global average for girls.
UNICEF's 2024 data on child marriage: twelve million girls per year are married in childhood globally; one in five young women aged twenty to twenty-four were married as children. Sub-Saharan Africa has the highest prevalence at thirty-one percent. The U.S. retains legal child marriage in thirty-four states; at least three hundred fifteen thousand minors were legally married between 2000 and 2021, eighty-six percent of them girls. (AoH Ch9.)
5. Racial disparities — and the gap in the literature
The CDC's NISVS lifetime rape baseline for adult women, stratified by race: American Indian / Alaska Native, thirty-four-point-one percent (highest of any group); Black, twenty-two percent; white, eighteen-point-eight percent; Latina, fourteen-point-six percent. Black survivors of sexual violence are forty percent less likely to have their cases prosecuted than white survivors after controlling for assault type, location, and relationship to perpetrator — the "Black woman believability gap" documented across criminal legal systems. (AoH Ch9, citing CDC NISVS and prosecution-disparity literature.)
For Native women and girls: more than four in five Native American and Alaska Native women have experienced sexual violence; ninety-six percent of perpetrators against Native women are non-Native; until the 2013 reauthorization of the Violence Against Women Act, tribal courts lacked criminal jurisdiction over non-Native offenders on tribal lands. As of June 2023, three-point-five percent of missing persons in the National Missing and Unidentified Persons System were identified as American Indian or Alaska Native — more than three times their share of the U.S. population. (DOJ Office of Justice Programs MMIP page; GAO MMIW report.) The Urban Indian Health Institute maintains its own independent missing-persons dataset because the federal databases cannot be trusted to count Native deaths and disappearances accurately.
Mainstream U.S. data sources do not carry tight, current, race-stratified child sexual abuse–specific lifetime prevalence figures comparable to the CDC's "one in four girls" headline. The strongest available substitutes are: the National Children's Alliance compilation of federal child-maltreatment data, which gives race-stratified per-1000 rates for maltreatment overall (AIAN children at fourteen-point-three per thousand, Black children at twelve-point-one per thousand); the CDC NISVS adult-women rates above; and adjacent scholarship in the adultification, MMIP, and child-welfare-disproportionality literatures.
The race-stratified CSA-specific lifetime prevalence number — the clean comparable answer to what percentage of Black girls in the United States are sexually abused before age 18 — is a gap in the literature itself. Researchers have called for it. The federal data infrastructure has not delivered it.
The strongest adjacent scholarship lives in four places:
- Adultification. Georgetown Law's Center on Gender Justice & Opportunity, Girlhood Interrupted: The Erasure of Black Girls' Childhood (2017), 2025 bench card update. Finding: adults view Black girls as less innocent and more adult-like than white peers, including girls as young as five to nine. When violated, Black girls are persecuted and labeled facilitators of their own abuse. (georgetown.edu; 2025 bench card.)
- The believability gap. African American Policy Forum / Kimberlé Crenshaw, "Say Her Name: Resisting Police Brutality Against Black Women." The canonical citation for the broader believability and visibility gap.
- MMIP. DOJ Office of Justice Programs Missing and Murdered Indigenous Persons page; GAO Missing and Murdered Indigenous Women report (gao-22-104045). Federal-data layer for tribal communities.
- Child welfare disproportionality. National Children's Alliance compilation of federal Child Maltreatment data — the per-1000 maltreatment-rate breakdown by race (AIAN 14.3; Black 12.1) is the closest federal comparison number.
The reason to name the gap is that the gap is doing work. When mainstream coverage describes "the one in four girls" or "the ninety percent perpetrator-known" figures, those numbers are reporting the aggregate U.S. female population. Inside the aggregate, the rate for Native and Black girls is meaningfully higher and the institutional response is meaningfully worse. The federal data infrastructure has not been built to publish the comparable race-stratified CSA figure. So the disparity reads as anecdotal where the aggregate reads as evidence. The architecture of how the U.S. counts CSA is itself part of the architecture this pack is naming.
What this means for the conversation
For providers, parents, and public-health-curious readers asking is this really that common? — the answer is yes. One in four girls in the U.S. One in twenty boys. Ninety percent of perpetrators known to the child. Three-quarters of assaults unreported. Average disclosure age fifty-two. One in five survivors never disclose at all. Globally, one in five girls and one in seven boys, with the U.S. running worse than the global average for girls. Across five U.S. institutional cases — Catholic, USAG, BSA, Penn State, SBC — over ten billion dollars in aggregate liability, the same mechanism in five different uniforms.
The Epstein operation is unusual because the paper survived. The structure it documents is the modal U.S. case. The grooming, the recruitment pyramid, the institutional cover, the silencing, the delayed accountability — these are not unique. The mechanism is the same whether the perpetrator is a billionaire, a priest, a coach, a scoutmaster, a youth pastor, or a neighbor.
The conversation that follows this one — in the pack and in the book — is about what gets said and not said in the public response. The medical taxonomy debate. The legal age-of-consent gradient. The "barely legal" framing. The DSM-5 paraphilia decision. Those are not separate questions from the prevalence data. They are the permission structure operating inside the prevalence reality. Now you have the prevalence reality.