top of page

The New Yorker: Why Doesn’t the U.S. Have an Accurate Count of Child-Abuse Deaths?

By Dan Hurley

Tanner Bivins was declared dead on January 5, 2018, at a hospital in Paducah, Kentucky, the week before his sixth birthday. As the Louisville Courier-Journal reported, Tanner was emaciated and dehydrated, and his scalp was covered with sores. Authorities later found his home strewn with garbage and vermin. The house had no utilities and no heat; the temperature had dropped to eleven degrees the night that Tanner died. According to the National Child Abuse and Neglect Data System (N.C.A.N.D.S.), an arm of the Administration for Children and Families, an estimated 1,720 children in the United States died due to abuse and neglect in the fiscal year 2017, the most recent year for which national statistics are available. But a death like that of Tanner Bivins would not have been included among them, because the medical examiner’s autopsy lists Tanner’s cause of death as “undetermined.” (As a state review panel noted in 2017, “Neither child maltreatment nor homicide is typically entered as the cause of death on the death certificate . . . because investigations necessary to make those determinations have not been completed at the time the death certificate is filed.”) The “undetermined” finding means that Kentucky’s social-service agencies will not report Tanner’s death as being the result of maltreatment. In part because of such cases, it is virtually impossible to find anyone who is both well versed in national child-abuse statistics and has faith in their accuracy. “The numbers could easily be three times or four times higher than what’s reported,” Emily Putnam-Hornstein, the director of the Children’s Data Network in California, said. “Whatever’s being reported to [N.C.A.N.D.S.] is an undercount.” In 2016, Kentucky reported that fifteen children died of abuse or neglect, for a rate of 1.48 per hundred thousand. (The national average is 2.32 per hundred thousand, while two states next door to Kentucky, Indiana and West Virginia, report death rates of 4.96 and 4.87, respectively.) Standards for accounting for child-abuse deaths vary wildly from state to state. In some Ohio counties, a death is not counted if there are no surviving siblings in the home. In Michigan, in some cases, a death may not be counted if the assailant is not a parent or legal guardian—for example, if a mother’s boyfriend kills her child. Alaska has no set definition for what constitutes a maltreatment death, according to Jared Parrish, the senior maternal and child-health epidemiologist in Alaska’s Division of Public Health. “Our child-welfare agency does not investigate all child deaths,” he said. “They usually just default to what the medical examiner reports. We know our numbers are grossly underestimated.” In 2016, a congressional commission on eliminating fatalities due to child abuse and neglect recommended that a nationwide “uniform definition” for counting such deaths and life-threatening injuries be rapidly designed and implemented. “Unless you have good data, you can’t tell if anything you actually do works to lower the rate of these fatalities,” Rachel Berger, a Pittsburgh pediatrician who led the commission’s research team, said. “The idea that every state has a different definition of child maltreatment is crazy. We focussed on fatalities because it seemed like the most straightforward place to start, but, if being counted as a maltreatment death depends on where you live, we’ll never get anywhere.


bottom of page