Thank you for stopping by this blog. This article is written for survivors of childhood abuse. If this is your first time here, please read the introductory notes about this site first.
“People who were abused as children end up abusing their own children.” You’ve probably heard this claim at least once. The transmission rate was once said to run as high as 70 or even 90 percent※1, but more recent research has rejected those numbers※2.
So how did this “cycle of abuse” hypothesis come about in the first place? This article walks back through the history of research on the intergenerational transmission of child abuse in three stages, and looks at why the popular belief spread — and why it was eventually overturned.
Chapter 1: The Early Research — “90% Transmission” From Interviews With Perpetrators Alone
The first method used to study the transmission of abuse was interviews with parents who had become abusers. When asked, “Were you yourself abused as a child?”, many of these abusive parents answered “yes.”
One after another, early studies reported strikingly high transmission rates of 70% to 90%※1. Some even reported “100% — every parent in the study had a history of abuse.”
These sensational findings sent a shock through the United States at the time, and spread the “intergenerational cycle of child abuse” hypothesis far and wide. But this research method had a fundamental problem: it studied only the perpetrators of abuse, with no comparison to non-abusive parents.
Chapter 2: The Mid-Period Research — Following People With Histories of Abuse
As criticism of the early method mounted, the methodology shifted significantly. Rather than asking abusers about their past, researchers began following people with histories of abuse forward in time, to see what happened when they later became parents.
In one such study, about 300 pregnant women, selected at random, were asked during pregnancy, “Have you ever been abused in the past?” Roughly 50 answered yes. A year later, the follow-up survey found that 10 of those women had gone on to abuse their own children.
Of the 50 with a history of abuse, 10 went on to abuse — this is given only as an illustration of the concept. The actual transmission rate reported in Kaufman & Zigler (1987) is roughly 30% (±5%)※2.
So why did the early studies produce a “90% transmission rate”? Because the two designs started from entirely different points. The early researchers began with parents reported to authorities for abuse, and asked them retrospectively, “Were you yourself abused as a child?” Nearly 90% answered yes. But this method has a fatal flaw: people who were abused themselves yet did not abuse their own children — the remaining ~70% you would see in a prospective design — are simply absent from the dataset.
Starting from “abusive parents and tracing backward” yields about 90%. Starting from “people who were abused and tracking forward” yields about 30%. The gap between those numbers is a difference in methodology. You can be looking at the very same phenomenon, but the picture you draw depends entirely on where you choose to stand.
Chapter 3: Modern Research — Three-Generation Tracking
The mid-period studies, in turn, drew further criticism. Self-reports of “I was abused in the past” still leave the problem of memory distortion and subjectivity. That is what produced the modern method: three-generation longitudinal tracking.
This method begins by following children whose abuse has been confirmed as fact. When that child becomes a parent twenty years later, do they abuse? And when their own child in turn becomes a parent, does abuse appear in that generation? Researchers follow three generations to find out.
Because this design demands enormous time and resources, the number of such studies is small. But because it does not rely on the subjectivity of recalled testimony, the objectivity is far higher.
The interim findings to date again point to transmission rates of around 30%. As the methodology has evolved — from testimony-driven early research, through a testimony-and-fact mid-period, into the fact-driven modern era — the conclusion is converging on “a transmission rate of about 30 percent.”
Thirty percent is by no means a small figure. But at the very least, the old claim that “most people who were abused will go on to abuse their own children” turned out to be the product of serious flaws in research design — and not, as the field eventually concluded, fact.
Comments