The Universe of Discourse

A child is bitten by a dog every 0.07 seconds...
I read in the newspaper today that letter carriers were bitten by dogs 3,000 times last year. (Curiously, this is not a round number; it is exact.) The article then continued: "children ... are 900 times more likely to be bitten than letter carriers."

This is obviously nonsense, because suppose the post office employs half a million letter carriers. (The actual number is actually about half that, but we are doing a back-of-the-envelope estimate of plausibility.) Then the bite rate is six bites per thousand letter carriers per year, and if children are 900 times more likely to be bitten, they are getting bitten at a rate of 5,400 bites per thousand children per year, or 5.4 bites per child. Insert your own joke here, or use the prefabricated joke framework in the title of this article.

I wrote to the reporter, who attributed the claim to the Postal Bulletin 22258 of 7 May 2009. It does indeed appear there. I am trying to track down the ultimate source, but I suspect I will not get any farther. I have discovered that the "900 times" figure appears in the Post Office's annual announcements of Dog Bite Prevention Month as far back as 2004, but not as far back as 2002.

Meantime, what are the correct numbers?

The Centers for Disease Control and Prevention have a superb on-line database of injury data. It immediately delivers the correct numbers for dog bite rate among children:

Age	Number of injuries	Population	Rate per 100,000
0	2,302	4,257,020	54.08
1	7,100	4,182,171	169.77
2	10,049	4,110,458	244.47
3	10,355	4,111,354	251.86
4	9,920	4,063,122	244.15
5	7,915	4,031,709	196.32
6	8,829	4,089,126	215.91
7	6,404	3,935,663	162.72
8	8,464	3,891,755	217.48
9	8,090	3,901,375	207.36
10	7,388	3,927,298	188.11
11	6,501	4,010,171	162.11
12	7,640	4,074,587	187.49
13	5,876	4,108,962	142.99
14	4,720	4,193,291	112.56
15	5,477	4,264,883	128.42
16	4,379	4,334,265	101.03
17	4,459	4,414,523	101.01
Total	133,560	82,361,752	162.16

According to the USPS 2008 Annual Report, in 2008 the USPS employed 211,661 city delivery carriers and 68,900 full-time rural delivery carriers, a total of 280,561. Since these 280,561 carriers received 3,000 dog bites, the rate per 100,000 carriers per year is 1069.29 bites.

So the correct statistic is not that children are 900 times more likely than carriers to be bitten, but rather that carriers are 6.6 times as likely as children to be bitten, 5.6 times if you consider only children under 13. Incidentally, your toddler's chance of being bitten in the course of a year is only about a quarter of a percent, ceteris paribus.

Where did 900 come from? I have no idea.

There are 293 times as many children as there are letter carriers, and they received a total of 44.5 times as many bites. The "900" figure is all over the Internet, despite being utterly wrong. Even with extensive searching, I was not able to find this factoid in the brochures or reports of any other reputable organization, including the American Veterinary Medical Association, the American Academy of Pediatrics, the Centers for Disease Control and Prevention, or the Humane Society of the Uniited States. It appears to be the invention of the USPS.

Also in the same newspaper, the new Indian restaurant on Baltimore avenue was advertising that they "specialize in vegetarian and non-vegetarian food". It's just a cornucopia of stupidity today, isn't it?

[Other articles in category /math] permanent link