Dogs Bite, but Balloons and Slippers are More Dangerous

Recently (Sept 11, 2005) I attended the lecture launching an important new book , " Dogs Bite, But Balloons and Slippers are More Dangerous " by Janis Bradley. The author Janis Bradley works at San Francisco SPCA's Academy for Dog Trainers. Thus she is very familiar with the shelter scene and with issues of discarded dogs; and she also has a lot of intellectual cross-fertilization with Jean Donaldson who also works at the SF SPCA.

(When I first started to review this book I slightly miss-stated the book title as "Dogs Bite, But Balloons and Bedroom Slippers are More Dangerous" : it is actually "slippers" rather than "bedroom slippers" in the title. The very sweet-faced dog on the cover -- whose breed ingredients I would defy God Herself to guess at -- is chewing on a bedroom slipper, with balloons in the background. I probably said "bedroom slippers" because I am so very given to alliteration.)

This book is of major importance to anyone concerned with issues of dog bites -- and of special interest to anyone concerned with fighting against potential or existing breed discrimatory legislation or with potential for legal or insurance limitations on pet ownership (and yes, that could include kitties, because you know they too do occasionally bite or scratch with high probability of nasty infection).

We were told by publisher James & Kenneth (= Dr Ian Dunbar) that anyone who purchased a copy at this lecture would be GIVEN a second copy FREE if they promised to give said copy to anyone who is "in the media". (My free copy was given the next day to the host of a local radio and TV station here in Davis who is also an excellent web designer and a great cat person. She had had me as a guest on her radio show some weeks earlier to speak on the subject of , guess what, "Dog Rescue" ; and that broadcast is continuously available as a pod-cast on the web .)

The main thesis : serious dog bites are rare .

The crucial thesis of the book is that dogs are one of the least dangerous things in our lives.

The main message of the book is that , considering how many dogs there are and how much contact there is between dogs and humans (ie how much "at risk" time), the real truth about dog bites is that while normal dogs do often give other dogs and humans some "back off" messages by a hard look or growl or air-snap, dogs really do not bite very often (surprisingly seldom considering the many indignities that we heap on them) and that the vast vast vast majority of dog bites are at the band-aid level or below it, ie do not cause more than very trivial injury, with serious injuries being rare and fatalities being so very very extraordinarily rare that you have 5 times as much chance of being killed by lightening as being killed by a dog.

In the first several chapters, author Bradley carefully goes through the statistics and studies for a number of causes of accidental injury and death , from the obvious big common causes like cars , falls, fires, etc to the intermediate ones like guns, poison, drowning, bicycles, etc to the really rare causes like lightening, bedroom slippers, and choking on balloons. Dog bites are among the absolutely rarest causes of serious injury and death. Dog bites kill slightly fewer children (annually in the US) than does choking on balloons. Dog bites cause somewhat fewer injuries (annually in the US) than do falls caused by bedroom slippers. Hence the title of the book.

Also, for comparison, she cites the most shocking of the non-accidental cause of children's deaths : parental, familial, or caretaker abuse, which kills enormously more children than dogs do. (She might have done well to add that as a non-accidental cause of adult death, murder by one's spouse ranks shockingly high, vastly higher than death by dog.)

All of the statistics are illustrated by dramatically clear graphs, which for most of us will bring home the comparative incidence more clearly than would the numbers alone.

Author Bradley analyzes the most frequently cited published studies and points out the very serious flaws in many of them , eg poor sampling and extrapolation from tiny samples to guestimates of national figures, eg poor questioning of informants such as very vague questions. She also points out that results from surveys using vastly different sampling or questioning or criteria cannot be compared to one another. Most of the media hoopla about "an epidemic of dog bites" is based on such faulty surveys and unjustified extrapolations. There is in fact no evidence that the incidence of dog bites relative to amount of contact (opportunity for bites) has actually increased.

Likewise she brings to attention the few well done studies that have some relevance and reliability and which show rather stable levels of dog bites over the years. (And, by the way, there is a good study showing , amoung other findings, that letting your dog sleep on your bed with you has no effect for good or ill on the likelihood of aggressive behavior.)

It would have been nice to add some well documented statistics on the larger number of lives unquestionably saved by dogs, ranging from the family dog heroes who haul a kid out of the water or get the family out of the burning house to the professional or quasi-professional search and rescue dogs. Most likely no one has gathered these statistics, and certainly the media does not publicize them. The author does review the growing evidence of dogs (and other pets) as beneficial to the health of people who live with them. Medical science has only recently begun to take this beneficial aspect of dogs seriously and to start gathering numerical evidence.

Why are we so afraid of dog bites ; why do we over-react to them ?

Human perceptions of risk and consequent levels of fear are often out of proportion to the reality of relative risk. This is not just for dog bites but for many other risks. (Maybe most of us were not paying attention to our courses in Probability and Statistics. Maybe most people never took such a course.)

Bradley analyzes many of the reasons why rare events are so often perceived as more of a risk than common ones. Generally we tend to over-react when it's a risk we did not choose to expose ourselves to, when it's a situation where we feel we are not in control (eg thus more fear of flying on a professionally piloted professionally maintained airplane than of driving ourselves in our own car), when it is a risk or situation that is unfamiliar to us (thus non dog people have more fear of dog bites than dog people, and likewise horses seem much scarier to the non-horseperson). We over-react when it is children that seem to be at risk; there are good evolutionary reasons for that one ! Plus, for dogs, well, our brains are hard-wired by evolution to be afraid of snakes, spiders, and big-fanged predators. Finally the news media makes more of a splash over rare events, because the common ones are not really "new" enough to be news. Contrary to the old saying, "Dog bites Man" gets a lot more coverage than "Man bites Dog." And , alas, while "Dog kills Man" is over-whelmingly newsworthy, the next best thing to a space shuttle explosion or a terrorist attack, the everyday tragedy of "Man kills Dog" , that minimum of six million pets killed every year in US shelters, is almost totally ignored.

Why don't dogs hurt us more often ? They really don't want to hurt us.

This chapter is a good discussion of the domestication of the wolf into the dog, including the now famous Russian experiments on breeding foxes for tameness and friendliness. Dogs so seldom hurt us because we have spent at least 14,000 years selecting them to be easily socialized to us and to extend to us the same ability to quarrel without inflicting serious injury that they have towards each other. At several points in the book Bradley suggests that we may have already done all we can do to minimize harmful biting in most normal types of dogs.

Bradley points out that some breeds have more recently been bred for lessened sociability and heightened neophobia, "bite first and ask questions later". These breds tend to be described in their own breed literature as "aloof" or "reserved" or as "suspicious of strangers" or as being "guard" or "protection" dogs. (And of course timid fearful people, who are the last ones who should try to handle such a dog, will buy them "for protection".) Not a good idea in her opinion if minimizing incidence of serious bites is a real priority, and so likely it would be wiser for most of us not to choose to acquire a dog of such a breed. The much maligned Pit Bull and related types are , however , NOT amoung these low sociability breeds; and the fighting lines in particular have been bred for very high bite inhibition towards humans so that they do not bite handlers or referees who are actually touching them at times during a fight.

This is followed by an analysis of "what is aggression anyway?" with illustrations from her own experience of various situations in which a person or animal got hurt due to another's actions, not all of which would be properly categorized as aggression and some others would not be categorized as unjustifiable agression. The types of dog aggression that cause the majority of bites, though not necessarily the most severe bites, are generally based in fear.

This leads into good discussion of the abysmal state of anti-dog legislation and the fallacies anti-breed legislation rests on and likewise good discussion of the insurance industry's reactions and attempts to cash in on this.

Breed studies and breed discriminatory legislation.

The book includes very good discussion of the many problems underlying all the studies that purport to say something about breed frequencies in serious dog bite incidents. The main point is that, as we all probably already know, the vast majority of people cannot recognize even common breeds let alone uncommon ones. Anything long haired and black and tan is likely to be labled as a GSD or GSD mix. Anything short haired and broad headed and muscular looking gets labled as a Pit or a Pit mix, unless it is black and tan in which case it gets labled as a Rott or Rott mix.

Anyone can create a risky dog or an extremely dangerous one by just reversing all the good advice on how to raise a stable well socialized bite-inhibited puppy (as I might put it, by acting like "the Anti-Dunbar" !). Thus , if the breeds currently popular with such anti-social people were to be made unavailable, there is no shortage of other dogs of other breeds or generic largish mixed breed who could be "weaponized" by any thug who cares to do so. The smarter thugs may chose dogs of breeds other than those currently popular as dangerous image dogs for the same reason one would choose a concealed weapon rather than an openly displayed one.

Bradley also discusses how poorly written, vague, inappropriate, and hard to enforce most of the "dangerous dog" or "viscious dog" statutes currently on the books are, then goes on to discuss how more appropriate and well defined laws can be , and in rare instances have been, written.

(My own view is that we might do better to define and legislate against "Dangerous Owners" and that all hearings to consider whether a particular dog is dangerous should give even more consideration at to whether the owner is so ignorant or irresponsible that he should be branded "dangerous" and forbiddent the priviledge of owning or keeping a dog. )

Homeowner's Insurance issues :

The author suggests that the real reason the insurance companies are making such a fuss about breed bans is that they are really hoping to split coverage for dog incidents OUT of normal Homeowners / Renters Insurance altogether so they can sell a separate "Dog Insurance" policy just as they sell separate Car Insurance. I note that states that require one to have car insurance in order to be allowed to have a Driver's License and/or to register one's car have not been very successful in eliminating uninsured motorists from their roads -- have you noticed that too ? So I doubt very much that any law demanding that all dog owners carry Dog Insurance would be very effective.

a sidebar on finding an insurer that does NOT practice breed discrimination :

After her lecture, Bradley mentioned in passing that she has her Homeowners Insurance with State Farm which does NOT have any breed bans or discriminatory criterion but only declines to give coverage to those whose dogs have actually bitten someone. (Now it seems to me for an insurance company to refuse to cover or charge higher rates for those who have actually had prior bad incidents, eg as they do for car drivers with moving violations or DUI convictions, is perfectly rational and perfectly fair. At least it is fair if you assume that dog bites are mostly due to some kind of faulty conduct, some negligence, on the part of the owner or to some dangerous quality of that particular dog himself , and that therefore a past bite predicts a higher risk of future bites. See below however for a model that compares the insurance companies reactions to dog bites to their reaction to much larger losses from natural disasters that can be predicted to re-occur.)

So today I did a little telephone research , by phoning an agent for each of the major Home Owner insurance companies that have agents in my town. The results are pleasantly surprising :

State Farm indeed does NOT have any breed restrictions -- not unless you would count declining to insure wolf and wolf-hybrid as a breed restriction. They decline to insure anyone with wild animals, and consider wolves and hybrids to fall in that category. (I rather agree. I didn't ask about coyote-dog hybrids, but the same rationale should apply.) They will decline to write new policies for those who have a dog which has actually bitten someone. I failed to ask if that meant any reported bite, any insurance claim based on a bite, or any claim for a bite actually paid -- these can differ a great deal. They have no limits on cats except for those of wild species -- sorry Siggie & Roy !

Farmers Insurance likewise does NOT have any breed restrictions. They will not insure you if you have an animal on whom there has been a prior claim which resulted in a pay out. This particular agent was clued in enough to mention claims paid because a dog knocked someone down causing serious injury. (I agree : I suspect that there could well be more adult deaths per year from dogs knocking an older person down causing a broken hip causing post-operative pneumonia causing death than there are adult deaths from dog bites. Indeed I would be willing to bet on this.)

Allstate however DOES have a list of dog BREEDS they will NOT cover and that list is currently purebred or mix of : Akita, Boxer, Chow, Dobie, Pit Bull, Pressa Canario, Rottie, and wolf-hybrid. Interestingly the agent did mention wolf-hybrid but did not mention pure wolf. They also will not insure those with a dog who "displays a vicious temperament", but the agent had no idea how that might be defined.

AFLAC has a list of BANNED BREEDS that includes GSD, Dobie, Rott, and Pit. that may not be the full list. The agent I spoke to said this list was "based on claims the company has paid out" but of course that the underlying data was not available to customers. he sounded very self-rightous about all of this and projected a "take it or leave it attitude" and was not impressed that a customer might actually chose to take their business elsewhere because of this dog policy. I have no idea if AFLAC will continue to insure you if your pet duck or goose has goosed someone painfully enough to create a claim. I did not ask about farm animals or livestock.

One of my correspondents says that Liberty Mutual also does NOT have any breed bans, or at least that they did not several years ago when she inquired before taking a policy with them. I have not had chance to check this out.

So the upshot is that I would encourage ALL of you when renewal time comes around to question your own agent about whether your current carrier has any kind of list of prohibited breeds and if the carrier does have such a list, switch your business to one that does not, ie to State Farm or Farmers (or any other you can find). When you do so, make sure to TELL both the company you are leaving and the one to which you switch that the switch was totally motivated by presence or absence of any breed specific restrictions. Vote with your wallet !!!!

(Note : if you have a horse, either on premises or boarded elsewhere, you should also be asking some specific questions about coverage for injuries caused by your horse. Your horse is much more likely to seriously hurt someone than your dog is. But if you are a horseperson , you know that.)

Some further thoughts of my own about the rightfulness of an insurance company declining to continue coverage AFTER your dog has made a bite that causes a claim.
On the one hand, this does seem reasonable and therefore fair to me because I do tend to agree with the underlying assumption that most serious bites involve some owner negligence or wrongdoing such as having failed to socialize the dog or failing to control the dog's interactions with others. And I do agree that serious bites often involve a dog who is a lot riskier than the average, ie who does present a future risk. Of course I advise professional help from a veterinary behaviorist for these situations, because both the dog and the owner need to get help making changes. (Sadly, there do exist some dogs who are so risky that they cannot live in society safely and for these humane execution becomes a sad necessity ; I've done it a few times myself, but it makes a wound that never fully heals.)
But then on the other hand , insurance companies will pay out huge amounts to you if your house has been destroyed by a flood or fire or tornado or hurricane, or other natural disaster or "act of God" (who conveniently cannot be sued), then will make no objection (or even insist upon) to your re-building in exactly the same location , a location known to be subject to periodic events of the kind that caused the damage they have just paid for, and they will continue to insure you for that same kind of damage.
So why, even assuming that a dog who has bitten (seriously) once is more likely to bite (seriously) again in the future than a dog with no previous known bites (an assumption that I would consider has some validity, though a smart owner would now change training and management in ways that could greatly reduce this risk, perhaps to below the average risk for a dog with no history) now considered to be uninsurable when a house in an area subject to repeated flood, fire, tornado or hurricane is still insurable ??? Of course there is no federal disaster aid for dog bites , as there is for widespread flood, fire, etc disasters. But I bet if you asked any insurance company if they would rather pay out all the claims in the next year for dog bites or all the claims in the next year for hurricanes, they would rather pay for dog bites.
When I look at my own insurance premium, which is for a Renter's policy so that the costs of re-building the house are not covered but only losses to my own personal property, I see that over 3/4 of my premium is for loss or damage to my personal property and less than 1/4 is for negligence liability , the latter including many many potential loses having nothing to do with dogs. (And I think that when I took out that policy I still had my horse, who represents more risk than my dogs do.)
Finally of course, let's say that your dog bites someone and they file a claim and you obey your insurance company's demand that you get rid of that dog (which usually means killing that dog) so they will continue to cover you. (You do it because your mortgage company demands that coverage be continued, unless of course you can pay off the entire bazillion dollars of remaining debt.) But let's say you have learned nothing from the bite incident about good dog choice, good dog upbringing, good dog management, nor any other risk reduction factors. So a while later you get another dog and you repeat all the same old mistakes. Has the insurance company's mandated killing of your old dog really reduced their exposure ? Wouldn't it make more sense if they instead mandated that you and the dog attend some kind of Dog-owners Education or some kind of behavior therapy program, as well as hiking up your premium rates for the next few years. This would be more like the approach for car insurance for drivers with an at fault accident , etc.

The more one thinks about it, the more irrational the insurance companies' attitude towards dogs seems. But then this would not be possible were it not for the irrationality of society's attitudes.

What can we do to further reduce the incidence of serious dog bites ?

At several points Bradley suggests that we may already have done all we can do genetically to most dogs to reduce incidence of biting. She also suggests that because our current system produces so very few serious bite, we might do well to be cautious about changing it because the changes may introduce other more serious dangers that have not been forseen. For example, making some breeds illegal and driving them "underground" certainly risks that these dogs will not receive their Rabies vaccinations, thus raising the incidence of human cases of Rabies above our current hard-achieved level of zero. For another example, reducing dog ownership in homes with children will almost certainly increase the numbers of children who become seriously asthmatic (growing up in a home with dog and/or cat greatly reduces incidence of asthma), thus causing far more suffering and death than that currently due to dog bites.

The most reasonable steps to reduce risks of serious dog bites are those any astute dog person would already know about. Raise puppies with tons of socialization and with careful attention to bite inhibition (teaching soft mouth towards human flesh), just as our good Dr Dunbar and other writers and lecturers have been telling us for so many years now. Educate parents more about supervising children's interactions with dogs and about how to "read" a dog's body language, and have these parents educate their children about how to behave sensibly and considerately around dogs. Finally and obviously, of course do NOT breed from any dog who is unreasonably aggressive or bad tempered or too hard to socialize ; don't breed from him even if he wins at Westminster. None of these suggestions are really news to any thoughtful dog person, and all are worth doing whether or not they further reduce the already low rate of serious bites, because all of them tend to make the dog to human relationship flow more pleasantly for both sides.

Odds and ends :

Somewhere in the book there is a reference to a study that confirms my long held belief that most of the dog bite fatalities involving children also involve a parent or caretaker of that child who was seriously inattentive or negligent in exposing the child to a known (known to theat parent) serious risk. Some of the news stories make parental malfeasance obvious, yet it is rarely or never directly discussed and criticised the way it would be if the child had died by falling into a nearby pond or other obvious inanimate hazzard. The reference is K Delise, "Fatal Dog Attacks : the Stories Behind the Statistics" , Anubis Press, NY. I have not yet had chance to look at this , but Bradley told me that it does show that parental or caretaker fault is a frequent feature of these cases.

This book has an extensive list of references, including many professional level literature published in peer reviewed journals for health or for animal behavior, and others from government sources such as Center for Disease Control. There are also references to local legislation and to popular press articles.

While there are a few areas in this book where one could make additional comments, these would mostly be footnotes to expand on the material.

There is one typographical error near the bottom of page 46 in which it says "the average treated dog bite is rated at the lowest level (of severity), 1 out of 6" and on the next page the material makes clear that the "severity of injury" scale actually goes from 1 to 16, where 16 is "serious to critical" (very life threatening). 1 means "fast healing with no lasting imairment" ; levels 1 to 3 are considered "minor". Over 92% of all treated bites are rated 0 (zero) on this standardized scale (as judged by the treating medical person), 7.5 % rated as level 1, leaving less than 1% rated level 2 and above.

I only found one real miss-statement that I caught and it is not one that actually changes any of the arguments : Bradley refers to "Mendel's work with fruit flies". Now I would hope that those of you who took Genetics 101 or read any good basic genetics text would remember that Gregor Mendel worked with garden peas, not fruit flies. It was Thomas Hunt Morgan and his students who really made Drosophila melanogaster the premiere species for genetics work, including a proof that the genes (at this point still an invisible theoretical entity) have physical residence on the chromosomes (visible under the microscope).   

(When I was a graduate student in Genetics at University of Wisconsin, I had a sign over my desk that read "The only good fly is one that unzips !" This was considered by some to be unspeakably rude and crude. So much for the liberality of the universities of the late 1960's early 1970's.)

Concluding thoughts :

Seriously, folks, this book is one that ALL dog people should read and be prepared to recite key facts from at the drop of an inhibition. We should ALL be discussing this book with (and perhaps giving a copy of it to) our insurance agents and with any local legislators (city councilpersons and so on) or state legislators whose ear we can catch (and pinch if necessary). we need to be educating the general public that , yes of course often dog bites can and should be prevented by good socialization and so on, but that there is NO dog bite crisis and the typical dog bite is a disinfect and bandage type of minor injury. DOGS ARE ONE OF THE SAFEST THINGS IN OUR LIVES as well as ONE OF THE MOST BENEFICIAL.

About the publisher :

This book is published by James & Kenneth <www.jamesandkenneth.com> in Berkeley , CA, which is the publisher of several of Dr Ian Dunbar's best books and also Jean Donaldson's "Culture Clash" and Pam Reid's "Excel-erated Learning". J&K is actually run by Dr Dunbar, which I guess should not come as much of a surprise. He was at the talk to introduce Janis Bradley and present her with the first copy off the press which was enclosed in a box that also contained a pair of bedroom slippers and was attached to a set of about half a dozen helium filled balloons. A very classy event overall.

(Incidentally, there is a new edition out of Donaldson's Culture Clash, which contains some added material and some issues on which her views have changed somewhat over the years. (The ability and willingness to modify one's views in the light of new evidence is the mark of real intelligence and intellectual honesty. What a pity so few of our elected governmental leaders have this quality !) The new issue integrates the key findings of "Dogs Bite but" into appropriate places in the text.)

J&K offers a substantial discount *** 55% DISCOUNT ! *** to ANYONE who buys in bulk for their books and videos. Originally they had dog trainers in mind, especially those who wanted to use one of these books as an assigned text, however it would be equally applicable to breeders who might wish to assign a text to potential puppy buyers (as I have urged for years) or give a book as a sort of "user's manual" to puppy buyers . The discount is available to ANYONE who is buying in the stipulated bulk.

Also the J&K site offers of Dr Dunbar's "Before You Get Your Puppy" and of his 16 pamphlets on various behavior topics. These download quite fast even over a dial-up. They are also available on Dr Dunbar's <www.siriuspup.com> and I think also on the Open Paw site, <www.openpaw.org> , Open Paw being yet another educational endeavour in which the good Dr Dunbar is heavily involved.

It seems clear that Dr Dunbar is a man with a mission , who puts his mind , his muscle (effort) , and his money where his mouth is.

A personal note :

Years and years ago, late 80s I think ( a few months after the Challenger disaster and its subsequent wide publicity) , during an Anti-PitBull Pogrom (brought on by a well publicized death of a very young child who, in the absence of adult supervision : Mom was absent and Dad in bed with the babysitter) wandered into a neighbor's yard and encountered the chained-up "Pit Bull" guarding said neighbor's pot patch, a dog widely regarded by the neighborhood as "dangerous") I wrote an article in which I pointed out in high satire style (J. Swift would have appreciated it !) that while dogs might kill a dozen people a year, "so what" : there were so many other things that killed a dozen every day -- guns, cars, smoking, alcohol, etc etc -- yet which we do not attempt to ban because we recognize that it is not these things which kill but rather the stupidity and/or misconduct of human beings using (mis-using) these things that is at fault. (Unlike Bradley , I did not research to find whatever reliable statistics might have been available at the time, relying instead on common knowledge and rhetoric.)

Of course today I would add that your chances of being killed by a terrorist , whether our own home-grown variety or any foreign labled variety, are far higher than your chances of being killed by a dog.