Note: FBI data on abductions by a stranger in a single year is 332.
Fact Checker: 58,000 children ‘abducted’ a year: yet another fishy statistic, March 17, 2015
By Glenn Kessler
A frustrated reader wrote The Fact Checker concerning a statement that had appeared during a NBC Nightly News report on Jan. 16: “According to the FBI’s latest study, more than 58,000 kids were abducted by non-relatives in one year.”
The reader had written to the network complaining about the use of the statistic, with copious research about its failings, but had never received a response. So he asked The Fact Checker to call attention to the misuse of this statistic, as we had recently done on such claims as women having only 10 percent of worldwide income or that 1,800 college students die a year from “alcohol-related causes.”
We don’t necessarily want to single out NBC. A quick search finds that in recent years that ABC News also improperly cited this statistic, though some additional context was included. It is also listed as a “key fact” by the
for Missing and Exploited
Children. National Center
Where does this statistic come from and how is it misused?
First of all, this is relatively old data. (New reports often get around this uncomfortable fact by often calling it “the latest data.”) The number is derived from data collected between 1997 and 1999, mainly from telephone survey in 1999 of 16,111 with adult caregivers and 5,015 youths. The results were then weighted based on U.S. Census data.
The 1999 results were released in 2002 by an arm of the Department of Justice in a series of reports, as part of the National Incidence Studies of Missing, Abducted, Runaway and Thrownaway Children (NISMART).
So, first off, you see this is an estimate, based on a survey, not based on actual incidents. NBC News claimed this was FBI data, which makes it sound like actual reports, but instead this was simply derived from a study conducted by academicians for the Justice Department.
David Finkelhor, a
professor who is director of the Crimes Against Children Research Center and a
key author of the study, readily said that the media frequently misunderstand
and misreport this figure. “I don’t think it is a reliable estimate and I don’t
use it very much,” he said. “I am really not crazy about any of these numbers.” University
of New Hampshire
The DOJ reports provide ample warning that the data should be handled with care, though sometimes you have to look hard at the footnotes. Two key components of the 58,000 figure are children who are reported missing by their caretakers or children who were missed by the caretaker for at least an hour but no report was filed. (Note: The “reporting missing” number is a subset of the “caretaker missing” figure.)
But the report warns that “both of these numerical estimates are quite imprecise and could actually be quite a bit smaller or larger because they are based on very small numbers of cases.” One of the footnotes is emphatic: “Estimate is based on an extremely small sample of cases; therefore, its precision and confidence interval are unreliable.”
As an example, the number of children reported missing is estimated at 12,100—but the possible range (within a 95 percent confidence interval) is between less than 100 and 31,000. Similarly, the number of children whom a caretaker missed at some point is estimated at 33,000, with a possible range of 2,000 to 64,000.
Finally, the overall figure of 58,000 also includes children whom the caretaker did not miss at all. The report gives numerous examples: a 17-year-old girl on a date who was detained by force and sexually assaulted; a 15-year-old lured into a school bathroom and assaulted; a 17-year-old detained and assaulted in a parking lot at a high school football game.
Either the children did not tell their caregivers, or they did and the caregivers never filed a report, sometime because they saw no injury or did not believe the child.
Here, the range is 24,100 to 92,400. The difference from the two earlier ranges is mostly a function of the sample size, Finkelhor said. “There is not much we can say about these figures with precision,” he said.
These incidents may meet a legal definition of “abduction” but generally do not conform to the public’s understanding of an abduction, as many also might be considered assault. In news reports, the number is often cited in conjunction with a very distinct crime of kidnapping. (The ABC News report, for instance, was pegged to the case of an 8-year-old who was killed by a stranger after getting lost on the way home from camp.)
In an effort to highlight the distinction, Finelhor said the report included an estimate for the number of “stereotypical kidnapping victims” in a year. (Such crimes are “abductions perpetrated by a stranger or slight acquaintance and involving a child who was transported 50 or more miles, detained overnight, held for ransom or with the intent to keep the child permanently, or killed.”)
The number of stereotypical kidnappings is significantly smaller: 115.
We should also note that the most recent (2014) FBI data on abductions by a stranger in a single year is 332, which is presumably the figure that NBC News should have used, since it claimed to have cited FBI data.
The Pinocchio Test
Finkelhor noted that the Denver Post in 1986 won the Pulitzer Prize for Public Service for exposing the dubious and exaggerated statistics associated with missing children. Yet, here we are, nearly 30 years later, with the media still misusing the data in a frightening and unhelpful fashion.
The DOJ studies, as written, are fairly clear about the limitations of the data, though perhaps the update, when it is released in 2015, could more clearly highlight the caveats in bold type. Sometimes reporters need to be hit over the heads with a two-by-four.
In the meantime, Four Pinocchios to any news organization that lazily cites this statistic without noting its limitations — and also how old it is. The issue of missing and abused children is too serious to be waylaid by fishy statistics.