[Update 6:31 pm: The Denver Post has issued a revised correction for the online article in question.]
[Update December 29: Joey Bunch related that he takes responsibility for the mistake and apologizes for his initial reaction. For my part, I am satisfied with the way the Post has handled the issue.]
In their article for today’s Denver Post, Joey Bunch and Kieran Nicholson claim, “More than 500 children in the United States die in gun accidents each year, according to the U.S. Centers for Disease Control and Prevention in a 2007 report, which estimated 1.7 million children live in homes where guns are kept.” However, there seems to be no factual basis for that claim.
As Bunch is listed as first author and his contact information appears below the article, I contacted him to see where he got his figures. Unfortunately, in a series of emails (see below) he flatly refused to provide me with a citation. Apparently that is because no such citation exists.
CDC provides a search page for reviewing mortality statistics. The results for unintentional firearm deaths for 2007, ages zero through seventeen, is 112. Notice that the anti-gun Brady Campaign reports comparable figures. (Of the estimated 2,436,652 deaths in the U.S. in 2009, a total of 588 for all age groups resulted from “accidental discharge of firearms.” Final figures for 2007 show a total of 613 deaths. Please see pages 19 and 39 of the linked CDC report, and notice that I provided an actual citation for my claim.) To get figures as high as Bunch claims, one has to look at decades-old data. (Note that, in this article, I am concerned only with Bunch’s factual claims. I will address the “big picture” issues elsewhere.)
So how did Bunch get from 112 to “more than 500?” I don’t really know, given he refused to tell me. I do have a guess, however. A top Google hit for “kids die guns” is a 2008 article from MomLogic. That article includes the same numbers as Bunch uses — “more than 500″ and “1.7 million households.” My guess is that Bunch cribbed these figures (from this web site or a comparable one) without bothering to verify them or even review their meaning.
Here’s what MomLogic has to say: “More than 500 children die annually from accidental gunshots. … Last year, a study conducted by the Centers for Disease Control and Prevention found more than 1.7 million children live in homes with loaded and unlocked guns.”
What is similar between this article and Bunch’s article is that both include the same year for the CDC claim (2007), both include the phrase “more than 500 children,” and both include the phrase, “1.7 million children live in homes.” One important detail to notice is that the MomLogic article does not cite the CDC for the “more than 500″ claim. Also notice the important qualifier in the MomLogic article about the 1.7 million households: these are “homes with loaded and unlocked guns.” Bunch offers no such qualifier, rendering his statement wildly inaccurate. (Neither MomLogic nor Bunch actually cite a specific CDC publication.)
I did find some support for the claim about 1.7 million households, but this comes not from CDC but from the American Academy of Pediatrics. (Perhaps there was some association between CDC and the Academy.) (Update: As USA Today relates, the authors of the study did have a direct relationship to the CDC.) That 2005 article states, “Findings indicate that ~1.69 million (95% confidence interval: 1.57-1.82 million) children and youth in the United States <18 years old are living with loaded and unlocked household firearms.” USA Today offered a popular summary of the study. However, the study is based on survey data, so its conclusions are suspect. (Please notice again my actual citations.)
At this point, then, the Denver Post either needs to come up with an actual citation supporting Bunch’s claim, or else issue a correction.
And, in general, I encourage reporters to a) actually have real citations backing up their claims (see also my write-up of a 2008 incident), and b) make those citations available to those who ask for them. Anything less constitutes journalistic negligence.
Following is today’s email exchange between Bunch and me:
Ari: Dear Mr. Bunch, You write: “More than 500 children in the United States die in gun accidents each year, according to the U.S. Centers for Disease Control and Prevention in a 2007 report.” Please send [me] your citation for that claim. Thanks, -Ari
Joey: CDC. I cited my source.
Ari: I see that you wrote down CDC in your article. The problem is that when I look at the CDC web page, I find very different numbers than the ones you claim. So what I’m asking you for is the actual citation for a specific document that backs up your statement. Please provide that, and stop being coy. Thanks, -Ari
Joey: It took me all of about 3 minutes to find that report. With all due respect, Ari, you’re a columnist for a competing newspaper, do your own work.
Ari: Joey, If you found it, then please *send me the cite*. The fact that I write for the Grand Junction Free Press (hardly a competitor to the Post) is entirely irrelevant. I did my own work, as I mentioned, and I found different figures. So now, again, I ask you to back up your claim with a specific citation. Thanks, -Ari
Joey: I told you the name of the report and the year it came out. Would you like me to print it out and drive it to your house? I’ll pick up coffee and doughnuts on the way. Good luck with your story, Ari.
Ari: No, I would would like you send me the link to the relevant document, or, if the document is not available online, the title and authors of the printed document. That will be trivially easy for you to accomplish, so please, again, send me the citation. Thanks, -Ari
Joey: I do freelance work sometimes. I’ll send you a bill for research, and when it’s paid I’ll spend my time doing your work. Failing that, you could call the CDC and ask them to send it for you. There could be a per-page fee for that, however. Have a nice day.
Ari: Dear Mr. Bunch, According to your own claims, you’ve *already done the work*, and it took you “all of about three minutes.” If you’ve already done the work, and found the citation that informs your article, then it will take you about ten seconds to send me the relevant link (or title with authors). As a writer for the Denver Post, you have a responsibility, both to your readers and to the owners and managers of the paper, to back up your factual claims with specific citations. Please do so at this time, and please stop acting so evasive and frankly unprofessional. Thanks, -Ari
Joey: One more time and the last time I’m saying it: do your own work. You work for a newspaper. You are a journalist. Do your own work. Conversation over.
Update: Apparently the conversation is not yet over. After I sent an email to several representatives of the Denver Post linking to this write-up, Bunch again responded, claiming (among other things), “I told you the name of the report.” I wrote back noting that he has not, in fact, provided me with the title of the report or anything like a verifiable citation. I will update this article when and if Bunch provides me with an actual citation to the alleged report in question.
Update: Kevin Dale, news director for the Denver Post, states via email, “We are correcting the statistics. Page 2 in tomorrow’s paper. We’ll be correcting the online story shortly. Thanks for bringing it to our attention. We take our accuracy very seriously indeed.”
Update 5:03 pm: I sent a follow-up email to Dale:
Dear Mr. Dale,
Thank you for promptly following through on the matter of the claimed gun statistics published in today’s Denver Post.
Unfortunately, the Denver Post’s online “correction” also is in error [as of the time of this update].
The “correction” states that in 2007, 138 children died due to “fatal shooting accidents.” But that figure is for ages 0 through 19. Last time I checked, the legal age of adulthood is 18. Therefore, the correct figure is for ages 0 through 17, which is 112 (as I mentioned in my write-up). (While the figures vary only slightly in this case, I still think the Post ought to get its basic facts straight.)
I invite you to see for yourself here:
Moreover, the online article continues to falsely state, “The CDC also estimated 1.7 million children live in homes where guns are kept.” The Post’s claim here is wildly inaccurate. The figure actually pertains to children “living with loaded and unlocked household firearms.” The number of children living in homes “where guns are kept” is many times that amount.
Again, I invite you to see for yourself here:
(Anyway, that article relies on survey data, which are notoriously unreliable in these matters.)
Thank you for your attention to this matter.
Update 6:31 pm: The Denver Post has issued a revised correction for the online article in question.