I’d like to clarify that when I warned 2 people about their own dogs I didn’t just walk in the door and see a pit bull and start sprinkling holy water on them in a knee-jerk fear of the dog. I observed the behavior of the dogs.
Pit Bull defenders here have claimed it isn’t even the owners, that even a Pit trained to be a Fighting dog is no more aggressive vs humans than any other dogs.
That’s not what I’ve been seeing. Any dog’s training (or abuse) can turn it into a danger to humans. By nature, pits are no more aggressive to people than other dogs, but every big dog is potentially dangerous. There’s no doubt they can and do hurt and kill people. The same is true for every large dog breed.
While that’s absolutely 100% true for instantly obvious reasons, there’s a strong distinction to be made between “Trained as a Fighting Dog (by someone who knows what that means in terms of controlled aggression)”, “Trained as a mean dog that fights (which is the vast majority of tough-guy pit bulls)”, and “Not trained effectively at all (which is the fair majority of most other pit bulls)”. As a society, I think since the days when Pits and Rottweilers were common family pets we’ve lost a lot of societal knowledge on how to handle dogs bred with that protective instinct and don’t-quit fighting instinct.
Honestly, I see it the same as a lot of other issues: If you were able to mandate serious, professional training for owners of the dangerous thing, the dangers would be mostly or completely mitigated except in the case of people who deliberately wanted to commit harm.
[ignorance-fighting suit on] While I hold no particular brief for Pit Bulls, I can’t sit by and watch basic statistics being violently abused in this way.
Absolute number of fatalities is insufficient data on which to judge the “deadliness” of a breed. You have to know how that number compares to the prevalence of that breed in the general dog population.
Here, for example, is a collection of studies that shows a chart of dog attacks by breed for the period 1982-2012. As you can see, the category “pit bulls” leads the list in absolute numbers of almost all types of attacks, including fatal ones.
But if you look at the data more thoughtfully, it’s more complicated than that. Pit bulls are responsible for way more deaths, but they’re also a way bigger part of the dog population than most of their “competitors”.
Consider, for example, that 233 deaths in this period are attributed to pit bulls, which make up 4.4% of the dog population. In the same period, Presa canarios accounted for 14 deaths despite being only 0.04% of the dog population, Huskies for 24 at 0.05%, Akitas for 8 at 0.06% of the population, and Chows for 7 at a mere 0.01% of the population. Rottweilers came in at 81 deaths and 2.2% of the dog population.
Look at what that says about the deaths-per-dog ratio of the different breeds. If we postulate a million dogs total, just for comparison’s sake, then the numbers shake down as follows:
**44000 pit bulls, 223 deaths: 0.0053 deaths per dog
22000 rottweilers, 81 deaths: 0.0037 deaths per dog
500 huskies, 24 deaths: 0.048 deaths per dog
400 presa canarios (bullmastiffs), 14 deaths: 0.035 deaths per dog
600 Akitas, 8 deaths: 0.013 deaths per dog
100 Chows, 7 deaths: 0.07 deaths per dog**
Yes, there are many more than a million dogs total so the absolute numbers on those ratios are wrong, but their relative sizes are the same no matter what the total population. According to these relative sizes, rottweilers are nearly as deadly as pit bulls, huskies are nearly ten times as deadly as pit bulls, and Chows are more than ten times as deadly as pit bulls.
Now, it behooves me to point out that when you’re talking numbers of deaths as small as 7 or 8, you can’t trust your sample as statistically representative of the population as a whole. But even limiting the comparison to numbers in the double and triple digits, it’s clear that pit bulls are nowhere near as outstanding in the “Deadliest Dog competition” as you’re trying to claim. Mostly, pit bulls kill a lot more people than their “competitor” breeds because there are a lot more of them.
Yeesh. [i-fs off]
It’s an interesting article, but I’m deeply suspicious of their conclusion that there are 440 pit bulls for every chow out there. At least in my region, this is totally inaccurate. It turns out that they’re basing their population statistics off of for-sale ads in classified papers. If a breed is less likely to be offered for sale, its population numbers will reflect that. Our shelter in NC saw a tremendous number of chow mixes come through in the 2000s, probably more than we saw pit bulls come through.
I don’t know if this way of accounting for breed has been analyzed, but I wanted to throw it out for consideration.
According to the latest AKC stats, pit bulls (APBT, AM STAF TER, STAFF BULL TER) are not even on the top ten list of registered dogs, while the Rott is #9:
http://www.petinsurance.com/healthzone/pet-articles/pet-breeds/Top-10-Dog-Breeds.aspx
Pits are not in the top 10 in 2010 or 2009 either:
That same link you referenced, closes with this advice (very last paragraph / summary):
“Pit bulls and Rottweilers are accordingly dogs who not only must be handled with
special precautions, but also must be regulated with special requirements appropriate to
the risk they may pose to the public and other animals, if they are to be kept at all.”
I believe you may be misreading that 4.4%; because in the notes, Clifton states:
“pit bull terriers and their close mixes make up only about 3.3% of the total U.S. dog
population, according to my frequent surveys of regionally balanced samples of
classified ads of dogs for sale, but they constitute 29% of the dog population in U.S.
animal shelters at any given time, according to my 2011 single-day shelter inventory survey”
PIT BULLS ARE NOT THE MOST POPULAR TYPE OF DOG IN THE USA; not even on the top ten.
This does not answer Kimstu’s problems with these stats, though. Do you agree or disagree that it’s fair to analyze stats by deaths per dog? That is, if 10 pit bulls kill 5 people, that’s worse than if 1000 poodles kill 6 people?
No, that’s from his cited research paper. See my post above about how that stat might be problematic, though.
But the study data that I cited is not based on AKC stats, which as you note pertain to registered purebred breeds.
I did not misread the 4.4% figure, as you can see for yourself from the data in the article. Perhaps the discrepancy is due to the fact that the attack statistics are given for the US and Canada combined, whereas the sentence you quoted refers specifically to the “U.S. dog population”.
(Moreover, even if we altered the pit bull population percentage figure to 3.3% instead of 4.4% in the previous comparison, that would still put the pit bull’s “deaths per dog” figure at only 0.0068. That’s still considerably lower than the ratio for most of the other breeds listed; even the rottweiler’s “deadliness ratio” wouldn’t drop below half of the pit bull’s.)
Goalpost moving. I never said that pit bulls were the most popular type of dog in the US, nor even in the top ten.
I simply pointed out that they are numerous enough that their high representation in absolute numbers of attacks doesn’t necessarily mean they’re way more dangerous than other high-attack-incidence breeds.
And so far, no intelligent statistical arguments or superior data have been put forth to rebut that observation.
Kind of undercuts your attempts to argue that pit bulls are so way way WAY more dangerous than Rottweilers, doesn’t it? :dubious:
Like I said, I hold no brief for pit bulls and do not object at all to reasoned and statistically sound arguments that they are more dangerous than most other domestic dogs. I just don’t like to see statistical arguments ignorantly or carelessly misused to make a specious point.
Seriously, are you good with how that article calculated percentage of breeds? Does the idea of pit bulls being almost 500 times more common than chow chows seem reasonable to you?
Didn’t mean to sound dissy, but that’s not actually a rebuttal of the point I was making. You are quite right to point out that the data may be unreliable and doesn’t match your personal experience, but that’s not the same thing as providing evidence that the data is actually wrong.
And even if the relative proportion of pit bulls and chows as given in the article is in fact inaccurate (and I have no idea one way or the other from my own personal experience), that doesn’t refute my general observation that you can’t validly assess the “deadliness” of a breed without comparing some kind of “deaths per dog” ratio, rather than just looking at absolute numbers of fatalities.
As for whether it’s way inaccurate to estimate breed population percentages by comparing breeders’ ads: I honestly don’t know. AFAIK it’s true that the vast majority of pet dogs in the US are acquired by their first owner from someone who bred them for sale, so I wouldn’t be surprised if the percentages of different breeds of dogs sold as puppies do pretty closely mirror their percentages in the total dog population. Whether the ads collected by this study’s author are reliably representative of such sales is a whole other question.
I emailed Clifton for clarification, and he replied:
Consider, for example, that 233 deaths in this period are attributed to pit bulls, which make up 4.4% of the dog population. In the same period, Presa canarios accounted for 14 deaths despite being only 0.04% of the dog population
Presa Canarios are a pit bull variant -- I don't see this meaning much of anything, except that if you mix pit bull with mastiff, you get a bigger dangerous dog.
>> According to these relative sizes, rottweilers are nearly as deadly as pit bulls
This is correct. Over the years I have been tracking the data, both pit bulls & Rottweilers rate at 11 times more likely to kill someone than the average dog.
>> huskies are nearly ten times as deadly as pit bulls, and Chows are more than ten times as deadly as pit bulls.
These breeds actually come in at four times more likely to kill someone than the average dog.
(There is an interesting stair-step phenomenon at work here -- lots of breeds clustered at four times more dangerous than average, none between four times & 11 times, & then pits, Rotts, & many pit variants at 11 times.)
>> Otherwise Clifton reports is being read as pits being the most popular dog at 4.4%
Large retrievers -- Labs, goldens, & their mixes -- are at 7.4%.
I see I have a typo in my current chart, showing only 1% for Labs. That will be fixed immediately -- Labs have never been nearly that low.
From ANIMAL PEOPLE, July/August 2012:
Dogs most often listed for sale or adoption
2010 2011 2012 Avg. Ppltn.
Large retrievers 8.7% 5.2% 8.2% 7.4% 5.2 m.
Pit bull class 4.1% 3.3% 4.6% 4.0% 2.8 m.
Small terriers 2.2% 2.1% 2.1% 2.1% 1.5 m.
Poodles 1.4% 2.0% 2.8% 2.1% 1.5 m.
Setter class 1.0% 2.3% 2.3% 1.9% 1.3 m.
German shepherds 1.7% 1.4% 2.3% 1.8% 1.3 m.
Spaniels 1.2% 2.0% 2.0% 1.7% 1.2 m.
Beagles 1.2% 1.2% 2.3% 1.6% 1.1 m.
Chihuahuas 2.4% 0.7% 0.4% 1.2% 0.8 m.
Merritt Clifton
Editor, ANIMAL PEOPLE
P.O. Box 960
Clinton, WA 98236
Clifton also warned me to not use AKC ratings, as very few owners register their dogs with the AKC
Unfortunately, his reply doesn’t seem to have provided any specific data or statistics to back up his claims about breed deadliness. I’m not saying he’s wrong about them, I’m just saying we have no way of knowing why he thinks he’s right.
Good advice.
Looks like the “pit bull class” is indeed a very popular type of pet dog in the US.
He also pointed out an error in your math:
44000 pit bulls, 223 deaths: 0.0053 deaths per dog
> 500 huskies, 24 deaths: 0.048 deaths per dog
That should be 5,000 huskies, not 500.
Keep in mind, Clifton notes that USA shelter populations avg 29% pit bulls; also, note that most shelters will not adopt out pit bulls, and those that can, rarely do.
No, actually, that’s not an error in my math, that’s apparently another typo in his chart.
As you can see in the table on the page I linked to, huskies are listed as 0.05% of the dog population. And that works out to 500 huskies in a notional total population of 1 million dogs, just as I said.
:dubious: AFAICT, this implies that:
-
shelter dogs include a disproportionate number of pit bulls,
-
most of those shelter pit bulls aren’t available for sale or adoption,
-
but pit bulls nonetheless remain one of the biggest categories of dogs that are available for sale or adoption.
Consequently, it would seem that the “pit bull class” is a pretty popular type of dog. Just as I said.
What does “pit bull variant” mean? This is such a non-specific term, the author could include just about any molossoid breed he cares to. Breed a Presa Canario with a Boston Terrier (another molossoid) or a Pug (yet another molossoid) and see if you get a bigger, more dangerous dog. What a nonsensical statement.
Agreed on both paragraphs. I’m not refuting the data, and your observation about how to interpret it is correct; I’m simply casting doubt on it.
Well, the best stats I can find say that 20% of dogs are adopted from shelters. You might be trying to account for that with the word “first,” but I’m also finding that 35% of owners don’t get their pets spayed/neutered, which means that a lot of dogs are acquired by their first owner via the miracle of birth. And plenty of people get dogs at flea markets and from friends.
Edit: another source uses another means of determining dog numbers: AKC records. Also flawed, but interestingly, the proportions of pits to chow chows are wayyyy different here, about 60% as many chows as pits.