Skip to main content

There is a graphic making the rounds in child welfare that’s gotten a fair amount of attention lately. The graphic tracks the rate at which caseworkers “substantiated” various forms of child abuse and neglect allegations from 1990 to 2018.

It was created by Prof. David Finkelhor and his colleagues at the University of New Hampshire. In 2020, it turned up in a conference presentation from Prevent Child Abuse America. Then Chapin Hall at the University of Chicago used it in a PowerPoint presentation of its own. And then John Kelly raised it with Chapin Hall Executive Director Bryan Samuels on The Imprint’s weekly podcast.

In 1992, the earliest year for which I could find figures, about 2.9 million children were subjected to child abuse investigations.

You’ll probably find it striking too:

“I was wondering,” one advocate said to me upon seeing the graphic in Chapin Hall’s presentation, “does slide #7 look right to you?”

No, it doesn’t look right. Because it isn’t right. Some of the numbers in the slide aren’t real. 

In fact, about two-thirds of the numbers are made up. Finkelhor & Co. didn’t choose the numbers at random — but the numbers still aren’t real. They decided to compare the actual rate of “substantiated” neglect to twice the actual rate for physical abuse and three times the rate of sexual abuse. Below the chart Finkelhor and his colleagues say the doubling and tripling of actual figures for physical and sexual abuse rates was done to make trends in all three categories more visible in the same chart. Chapin Hall included this when it reproduced the graphic. Prevent Child Abuse America eliminated most of this explanation, leaving only the notation “x2” and “x3” next to the numbers for physical and sexual abuse.

The end result is a graphic that leaves the impression that there once was just as much physical and sexual abuse as there was neglect — and all three existed in massive proportions — until the child welfare system got the abuse figures to plummet.

But now, let’s do the same comparison without making any numbers up:

Scroll to Continue

Recommended Articles


The real numbers reveal that, through most of its modern history, the family policing system always has been obsessed with neglect (which, as I hope we all know by now, often means poverty). Using only the actual numbers also reveals that, though there really has been a significant decline in “substantiated” allegations of physical and sexual abuse — a decline that is clearly visible even without making any numbers up — the decline is from a very low number to an even lower number. 

No matter which graphic you use, they don’t support Bryan Samuels’ answer when Kelly asked him why abuse rates had fallen sharply but neglect had not. Samuels hypothesized that the decline in abuse could partially be attributed to the fact that what the group known as Children’s Rights calls “the child welfare surveillance state” has, in effect, scared would-be abusers straight and people are less inclined to abuse children (but, apparently just as inclined to neglect them) because they know they could be caught, lose their kids, and possibly be arrested. 

He is certainly right that we have built a massive surveillance system, one that nationwide will knock on the door of more than one-third of America’s children. In some big cities, almost every Black child will be forced to endure a child abuse investigation. But a closer look at the data calls his explanation into question. More than 60% of the decline in physical abuse substantiations and nearly 70% of the decline in sexual abuse substantiations took place between 1990 and 2000. So if Samuels is right, there should have been a gigantic increase in family surveillance during those years. Right?

There wasn’t.

By the 1990s, the child welfare surveillance state was already going like gangbusters. In 1992, the earliest year for which I could find figures, about 2.9 million children were subjected to child abuse investigations. By 2000, it was still about 2.9 million. So substantiated allegations of physical and sexual abuse declined sharply with no change at all in surveillance. In succeeding years the decline was much slower, and there has been almost no change since 2005. Yet during those years, the surveillance was indeed ratcheted up. By 2017, roughly 4.3 million children were subjected to child abuse investigations. So a more plausible interpretation is that when surveillance is ratcheted up, caseworkers spend even more time chasing down false reports and poverty cases and have less time to find the few children in real danger.

That’s because, as the unembellished numbers make clear, child abuse is rare. In fact, it’s even less common than these data suggest, since even definitions of abuse are so broad they can include many cases that are nothing like the horrors that come to mind when we hear the words “child abuse.” Precisely because the horrors are as rare as they are tragic, a giant surveillance state is the wrong way to try to stop them.


In his podcast interview with The Imprint, Samuels acknowledged that the child welfare surveillance state has a downside: that “sometimes there are families that are coming to the attention of child welfare not because they abuse their children, but because they’re poor, their children appear to be neglected, and people think they’re obligated to call the child welfare agency.”

The real, unembellished data suggest that the downside is the only side.

Richard Wexler

The Imprint