What is the positive dog trainer community’s stance on Breed Specific Legislation,
AKA banning particular dog breeds, deeming them too dangerous to allow people to keep them as pets? (or only allow people to keep them under special restrictions, such as “must always be leashed and wearing a muzzle in public”)
The above question was posted in a Facebook dog training group I’m a member of, and it soon attracted a flood of comments, which all condemned Breed Specific Legislation with a passion.
Some of the comments linked to an article by Zazie Todd in Companion Animal Psychology, titled Breed Specific Legislation Had No Effect on Dog Bites in Odense, Denmark. I’ve read that article, and this is my response to the article, and to the simplistic black & white thinking patterns filling the BSL debate.
(I was tempted to comment on the post – but would have been the only one that didn’t agree with everyone else – and after envisioning the shit storm that would hit me in that forum, I decided to publish this blog post instead)
My objective is to encourage nuanced thinking in the debate. My stance is that I don’t have a categorical stance to all BSL, “the devil is in the detail”.
“Breed Specific Legislation Had No Effect on Dog Bites in Odense, Denmark”
The article by Zazie Todd summarises:
This study joins a number of others in finding that Breed Specific Legislation does not work.
That conclusion is based on a new study, “The effect of breed-specific dog legislation on hospital treated dog bites in Odense, Denmark—A time series intervention study” by Finn Nilson , John Damsager, Jens Lauritsen, and Carl Bonander, which “used a time series intervention method on a detailed dataset from Odense University Hospital, Denmark, regarding dog bite injuries presented to the emergency department”. Their study found that:
The results indicate that banning certain breeds has a highly limited effect on the overall levels of dog bite injuries, and that an enforcement of the usage of muzzle and leash in public places for these breeds also has a limited effect.
That is a dramatic result, because Denmark banned 13 breeds in 2010 … and dogs from two of these breeds even had to be euthanised… yet bite statistics hadn’t improved 4.5 years later, according to the article, based on hospital bite statistics from Denmark’s third largest city.
13! If that isn’t evidence that “BSL doesn’t work”, then I don’t know what is!
Except if you look into the details, then it isn’t that black & white:
Pitbull Terriers and Tosa Inus had already been banned for many years when the BSL of 2010 was introduced, so there likely weren’t any in the country
According to Zazie Todd’s article:
In 2010, Denmark banned the ownership, breeding and import of 13 breeds of dog, including the American Staffordshire Terrier, the Dogo Argentino, Fila Brasiliero and American Bulldog. Two of those breeds, the Pitbull Terrier and the Tosa Inu, had to be euthanized. Any existing pets of the remaining 11 breeds could be kept, but they had to be muzzled and leashed in public.
Finn Nilson et al’s study says the same:
In this study, we focus on the case of a breed-specific legislation introduced on June 1st 2010 in Denmark (Hundeloven, §1a and §1b), banning the breeding, import and new ownership of thirteen breeds, specifically identified as dangerous. For two of these breeds (Pitbull terrier and Tosa Inu), all existing dogs were even ordered to be euthanized. For the remaining eleven (American Staffordshire Terrier, Fila Brasiliero, Dogo Argentino, American Bulldog,Boerboel, Kangal, Central Asian Ovtcharka, Caucasian Ovtcharka, Tornjak and Sarplaninac), an intermittent law was imposed on existing dogs meaning that these were forced to wear a muzzle and be on a leash in public places at all times
What a tragedie! All those Pitbull Terriers and Tosa Inus that had to be euthanised in 2010 because of the new law, their families, all the grief…
Except… in 2010, Pitt Bull Terriers and Tosa Inu had already been banned since the BSL of 1991, so technically there shouldn’t be any left in the country (although technically there could be, if they were born before the 1991 law was enacted, as the 1991 law didn’t require euthanasia, but they’d have to be at least 19 years old by 2010).
Let me guess that probably no Pitbull Terriers or Tosa Inus were euthanised as a result of the new law, although I don’t know for sure.
Also: if there were no Pitbull Terriers or Tosa Inus before the BSL of 2010, then banning them, and even requiring them to be euthanised, can’t be expected to impact on bite statistics, hence a conclusion can’t be drawn on that basis about whether the BSL “worked” for those two breeds.
Most of the banned breeds were uncommon in Denmark already before 2010
The banned breeds are:
- Pitbull terrier (since 1991)
- Tosa inu (since 1991)
- Amerikansk staffordshire terrier
- Fila brasileiro
- Dogo argentino
- Amerikansk bulldog
- Centralasiatisk ovtcharka
- Kaukasisk ovtcharka
- Sydrussisk ovtcharka
As mentioned, the Pitbull terrier and the Tosa Inu were already out of the picture long before the BSL of 2010 was introduced, so let’s look at the rest:
Livestock Guardian breeds
Almost half of all the breeds on the list are giant South/Eastern European or Asian Livestock Guardian Dogs (LGDs). They’ve never been common in Denmark, neither as pets nor as working LGDs.
These Livestock Guardian breeds are powerful, impressive, and invaluable in their natural work environment. They are generally independent, watchful, suspicious of strangers, and designed to roam vast remote outdoor spaces, guarding livestock such as as sheep against large predators, 24/7. LGDs in general aren’t great pet material in most cases, but the banned ones are those considered most risky in the hands of most pet owners.
Dogs of these breeds would also likely experience a lot of frustration as pets in Denmark, due to the restrictions of pet life and lack of space. Denmark is a densely populated country with a culture low in tolerance to risks and inconvenience, and typical pet life restrictions may include being expected to deal with strangers in crowded spaces, and not being allowed to bark at night time.
None of these breeds are “bad dogs” – on the contrary, they’re brilliant in the job they are bred to do – they just aren’t generally cut out for pet life in a restricted and space-limited environment, and are perceived to pose an unacceptable risk in society in that role.
Fila Brasileiro is another breed that can potentially be very dangerous in the wrong (pet owner) hands and environment; a very effective large guard dog that’s highly suspicious of strangers, and hard to socialise… AND they look irresistably cute, especially as puppies (try googling “Fila Brasileiro puppy”, and you will see what I mean).
I’ve never seen any of them Denmark, but according to the “Evaluering af hundelovens forbudsordning” report, 11 were being registered in Denmark in 2008-09, and none in 2011-12. Who knows if there were any in Odense.
A large muscular breed developed for large game hunting, descended from the Cordoba Fighting Dog, with a range of large breeds crossed in. The Dogo Argentino is generally described as a challenging breed “only for very experienced owners”, and highly selective with other dogs. It is easy to see how this breed could be dangerous to people and especially dogs in the wrong hands and places, like dog parks.
49 were being registered in Denmark in 2008-09, and 1 in 2011-12.
The South African mastiff is a giant guard dog breed that’s bred to be suspicious of strangers. 172 were being registered in 2008-09, 5 in 2011-12.
My point with mentioning the registration numbers is, that since most of the breeds were uncommon already before the ban in 2010, banning them would not likely influence bite statistics very much, if at all. It is also possible there weren’t any in Odense. Hence comparing the before-and-after bite stats in Odense doesn’t necessarily say much about the effect of BSL in other countries, where the same breeds may be more common.
… So why ban breeds that are uncommon anyway?
For example, looking at the banned LGDs:
The report “Betænkning om farlige hunde” by Justitsministeriet’s Udvalg om hunde (translates to something like “Report on dangerous dogs” by the Danish Ministry of Justice – describing part of the reasoning behind the law) describes the breeds that belong to “the group of shepherd- and guard dogs from South- and Eastern Europe and Asia” as
potentially the next generation of dangerous dogs.
In other words, they banned them because they imagined these breeds could potentially as pets in Danish society pose a serious hazard due to their size, strength, and natural dispositions, if they were to be commonly kept as pets, not because they had “done anything wrong”. (there also seems to be an assumption that these breeds would have pet appeal, which is probably not far off, seeing how they look like gorgeous gentle giant fluffy teddybears).
And – and this is a really important point:
It is obviously much easier to ban breeds before they become common as pet dogs, than after… Just look at how much controversy Pitbull bans cause in the US, where pitbulls & pitmixes are widespread and well established as family dogs.
Bans on popular breeds have tragic consequences for the dogs that “look wrong”, and for the people who love them, and the friends of friends of the people who love them, and people who just heard about it, and so on. It comes across like a form of state terror, and the personal stories trigger strong emotional public reactions – public outrage, divisiveness, sub-cultural trench wars.
Banning largely unknown breeds tends to go down much easier. I think that’s why the Danish BSL list of banned breeds has as many as 13 breeds on it, most of which most people had never heard of.
American Staffies and American Bulldogs
Now we get to the tricky ones, the “usual suspects” – the type of breeds that’s usually highly controversial in relation to BSL – in Denmark, that’s American Staffy (American Staffordshire Terrier), and American Bulldog.
(The Pitbull is only excluded because it had already been out of the picture for so many years in Denmark when the BSL of 2010 was introduced)
While not as common as in countries like Australia or the US, bull breeds are far more common in Denmark than all the other banned breeds. Below table from 2013, borrowed from the report “The 2013 report “Evaluering af hundelovens forbudsordning” (Danish – the title translates roughly to “Evaluation of the BSL”) shows the number of registered American Staffies and American Bulldogs (including their mixes, which is what “blanding” means) in 4 key areas of Denmark at the time of the report.
Jylland is the large peninsula that makes up most of the area of Denmark. Fyn is the large island between Jylland and Zealand. Copenhagen is the Capital and largest city of Denmark, and Aarhus is the second largest city of Denmark.
The bull breeds are known as popular, people-loving family dogs. They’re often playful, happy, easy going and friendly family pets, and have no other role – they are not misplaced working dogs, so if they can’t be pets, then they can’t be anything at all… and they have many fans.
I’m not sure exactly why these breeds are banned, but I can try some guesses.
Perhaps part of the reason for adding the American Staffy to the list of banned breeds, is that it looks so similar to Pitbulls, that it is difficult to distinguish between them, which is necessary (and errors can have major consequences) when one is banned, and the other is not. One could also argue that it is silly to ban one and not the other, when they are so alike in both looks and behaviour. Also, I’ve heard rumours from before the BSL, that people who wanted Pitbulls just imported them on fake papers, so they officially became American Staffies. I guess things like that may have irritated the law makers.
The American Bulldog is perhaps banned because it is perceived to be a similar type of breed.
The 2013 report that was commissioned to evaluate the effects of the BSL of 2010 (if any), says that “dogs of this type”, referring to American Staffies and American Bulldogs, are “commonly perceived to show a type of aggression closer to prey drive [hunting], than any other type of aggression”. The authors explains that while conventionally prey drive isn’t considered a form of aggression, they’ve chosen to include it as a form of aggression for the purpose of the report.
Attacks motivated by prey drive, and attacks motivated by classic forms of aggression, can look similar at a glance, but the typical duration, severity, and outcome differ. Classic forms of aggression (e.g. anxiety-driven aggression, resource guarding) aim to remove or deflect a threat – they are distance-increasing behaviours, whereas the objective of prey drive is to kill the victim, rather than make a threat go away.
Bull breeds have a reputation for propensity to dog-aggression, a “short escalation ladder” and insensitivity to other dogs’ calming signals when in a highly aroused state, and unwillingness to back off combined with high pain threshold, an affinity for hold-bites, and extraordinary perseverance when hold-biting. That’s also my personal impression (I’m now living in a country where they’re ubiquitous). All that is accentuated by a physique extremely well adapted to fighting – Herculean structure, thick strong neck, strong powerful jaws. Originally developed to grab & hold bulls, and later to win life & death dog fights, their ability to fight is no doubt superior to that of other breeds.
I personally feel that large bullbreeds do present a heightened risk to other pet dogs. That’s especially so when they’re owned by pet owners who are not motivated to manage them carefully, because they “don’t believe in the stereotypes”. Who take them to dog parks, walk them off-leash in the streets, and who don’t expect trouble and don’t notice warning signs. The ones whose dog “is good with other dogs” because the tail is wagging and it looks like he’s got a big smile on his face all the time. The average dog owner.
All that leads to the last caveat:
The BSL aims to prevent attacks on people and dogs, but the new study didn’t include veterinary bite statistics
There is in the wording of the Danish law about dogs made no distinction between the consequences of attacks on humans, and attacks on dogs. The legal consequence for a dog who mauls (skambider) an individual of either species, is euthanasia (regardless of the breed of the perpetrator). The list of injuries used to define mauling, applies to both humans and dogs.
I think the large bullbreeds are banned mainly due to the perceived risk they pose to dogs, and by extension grief, trauma and anxiety for pet owners of other dogs. Bite statistics from hospital admissions don’t measure that.
The Danish evaluation from 2013, “Evaluering af hundelovens forbudsordning”, did measure it. Data was collected from human hospitals and veterinary hospitals in order to compare bite statistics before and after introduction of the law, and the report found no change in bite statistics for humans, and a small reduction in some bite statistics for dogs, which the authors did not consider substantial.
Conclusion – The devil is in the detail
Bite stats are only as useful as the assumptions they rely on
Bite statistics are typically used to measure the effectiveness of BSL: when comparing bite statistics in the relevant area before VS after the BSL, did the number of bites decrease significantly after? If the numbers didn’t go down, then the BSL “didn’t work”.
The article “Breed Specific Legislation Had No Effect on Dog Bites in Odense, Denmark” and the study it quotes, concludes that “the BSL didn’t work” based on the fact that bite statistics from a hospital emergency department in Denmark’s third largest city had not improved 4.5 years after the 13 breeds were banned in Denmark. However, the reasoning is flawed because (summarising the caveats already described) it fails to consider that
- most of the banned breeds had never been common in Denmark
- two of the breeds on the list had been banned for many years already, and likely weren’t included in neither the “before” cohort, nor the “after” cohort
- the two banned breeds that are fairly common (although far from as common in many other countries), American Staffy and American Bulldog, are known primarily for dog-aggression, not people-aggression, but the new study did not include veterinary bite statistics. Danish dog legislation mentions attacks on people and dogs, but the new study ignored attacks on dogs
If breeds are not common prior to being banned – and/or if dogs of those breed didn’t bite anyone before the ban, more than dogs of other breeds, if the ban was merely preventive in regard to those breeds – then banning them can’t be expected to impact significantly on bite statistics. And if veterinary bite statistics aren’t included, then the result won’t be adequately relevant for dogs known for dog-aggression.
Bite stats can be influenced by many factors
Many factors can affect bite statistics beyond a breed ban, hereunder cultural aspects such as how pet owners train and manage their dogs and how people behave around them, risk awareness (which may increase after a BSL introduction and the public debates accompanying it), as well as technical statistical uncertainties (described in the 2018 Nilson/Damsager/Lauritsen/Bonander report), so bite stats isn’t a fool proof way to measure the effectiveness of BSL anyway.
Bite stats measure outcome, not risk or sense of safety
Objectives of BSL may include
- Lower the risk of severe dog attacks on humans
- Lower the risk of severe dog attacks on pet dogs
- Protect a general sense of everyday safety for people, hereunder dog owners
Risk is not the exact same thing as outcome. I can live with a high risk every day, but if I’m risk-aware and take sufficient precautions, the risk may never result in whatever bad outcomes it is that I’m risking.
However, even if the bad thing never happens, the risk may still influence my life negatively in the form of increased everyday restrictions, reduced opportunities, energy consuming risk management, and heightened stress and anxiety.
Different cultures have different baselines of risk tolerance (although of course individuals vary widely within cultures too). Denmark is a country with a low risk tolerance in general. There are many ways this shows and is justified. Crimes rates are low, there’s (almost) no dangerous wildlife, there’s a strong social security network (the “Welfare Society”), guns are illegal, etc. People are not used to living with a high level of perceived everyday risk to themselves, their friends, kids, and pets.
Taking dog ownership as an example: here in Australia where I live now (being originally from Denmark), I’ve experienced dog attacks on my dogs during walks and in other situations several times, and it is a risk I’m aware of and vigilant about on every walk. The risk is normal.
In Denmark, that risk wasn’t a thing I even considered, and dog attacks never happened to me or anyone I knew or heard of. I’m not saying it doesn’t happen in Denmark, but it seems like the risk is much lower, and I don’t think it is a risk most people would be willing to live with. It is taken for granted that dogs are fairly well behaved and safe around people and other animals. An example where that assumption is put into practice: pet dogs are allowed to travel on public transport in the cities (on a child’s ticket), which means dog owners are not nearly as dependent on having a car, as they are here in Australia.
The risk tolerance is much higher here in Australia, in general and amongst dog owners. I don’t think I could have my current dogs, who are dog-reactive and large, in Denmark (even if we ignore that one of them would probably be illegal based on the Danish BSL). Other dog owners would likely feel threatened by their reactivity rather than understanding (as most of them wouldn’t have been in a similar situation, or have known someone who was), and cause a lot of trouble for us. Low and high risk tolerance can be good and bad, depending whose shoes you’re in.
Perhaps some BSL is sensible
First a rhetorical question. We live in a globalised world… whatever animal is for sale and legal, anyone can import, breed, and sell to pet owners. Do the people who are categorically against BSL really want every conceivable breed to be available as a pet dog, no matter what purpose it has been selectively bred for?
Cute Fila Brasileiro puppies? Tosa Inu? Every breed of remote village-protecting, bear-fighting, wolf-tossing giant LGD?
I hope not.
Personally I’m relatively agnostic in regard to the large bullbreeds such as Pittbull, American Staffy, and American Bulldog. I do think it makes sense to either ban both Pitbull and American Staffy, or none of them, seeing how difficult and error-prone the process of telling them apart is.
I’ll also admit that I did feel a lot safer as a dog owner living in Denmark compared to Australia for many reasons, the majority of which has to do with other dog owners and the “canine culture” where I live here in Australia, in a typical outer suburb. Barrier aggression and dog reactivity is common, as are poor manners and ignorance. However, it also does play a role that large bullbreeds weren’t common anywhere I lived as a dog owner in Denmark (even long before the BSL), especially when combined with the above mentioned factors. I’m much less afraid of the ignorant Beagle owner’s dog, than the ignorant Pitbull owner’s dog.
On the other hand, considering the ramifications for the dogs and their owners, as well as how error-prone visual breed identification is, I don’t think it is a good idea to ban bullbreeds (Pit bull, American staffies…) in countries where they are common.
Let’s move on from the bullbreeds though.
There are other breeds whose typical traits I think are such an obvious hazard to people and pet dogs in the hands of clueless pet owners, and which are so generally unsuitable as pet dogs for their own good too, that I think it makes good sense to put those breeds out of reach of wanna-be owners.
I can’t understand the resistance to acknowledging that selective breeding has created breeds, originally (and sometimes still currently) developed for particular purposes far from and incompatible with modern pet life, whose combinations of traits make many of them high-risk in modern human society (although of course, pet life, and what pet owners need from their dogs, and general risk tolerance, differ from country to country… so the compatibility of breeds may vary from culture to culture).
Denmark is a densely populated country where people tend to live crowded and busy lives and be quite intolerant of unexpected everyday risks and inconveniences, so I think that’s particularly true of Denmark.
A showline Border Collie is no longer the full package of herding dog, because it has been bred for looks through generations, and some individuals may have lost almost all of their herding instinct and other natural inclinations that served the breed so well in its original role … However, some have not, and they may herd kids and bicyclists around. Herding breeds herd, guard dogs are suspicious of strangers, pointers point, labradors love retrieving and water. Not every individual is breed typical or have retained their original working traits, especially if we talk showlines, but it is definitely a thing.
Not to forget, many of the banned breeds are still being selectively bred for their working characteristics in the present day, for example the LGDs are still crucial for livestock protection in remote pastures in some countries.
Of course BSL could ban working lines but not show lines of certain breeds, but then telling them apart based on visual breed identification (or even breed DNA testing) would become near impossible.
The purpose of BSL isn’t to vilify breeds (although I admit that’s often a side effect), it is to prevent pet owners from taking on breed challenges that they discover too late they don’t have the skills to handle, and by then the repercussions of totally unnecessary tragedies are cascading through society.
Enforcement of BSL remains tricky and error-prone
Visual breed identification is difficult, error-prone, and unfair to people who buy a mixed breed puppy without much chance of figuring out what it will end up looking like as an adult.
Modern breed DNA testing has become much more reliable than it once was, at least when using the most advanced of the companies, and when detecting breeds that make up 25% or more of an individual.
I do understand some of the arguments pro visual breed identification – namely that the point of the BSL isn’t to remove particular breeds, but to remove particular types of dogs, whose mental and physical traits (e.g. large size and strength combined with suspicion to strangers, or dog-aggression) are perceived to be too big a hazard in society.
However, breed identification, when the consequences are serious like in the case of BSL, shouldn’t be left to some officers’ (who may not even have the necessary training) subjective judgement according to some breed standard-like metrics, that really aren’t all that relevant for whether those dogs are dangerous or not.
I think breed bans should be enforced only through a combination of visual breed identification, breed DNA test, temperament assessment, and assessment of the owner’s understanding of dog behaviour and ability to manage a dog of that type.
Idea for a better BSL framework
I’d prefer to see a rework of the BSL model, where instead of banning breeds outright (thereby indirectly labelling them “bad” in the minds of simplistically thinking people), “risky” breeds should be listed in a “restricted” category, where people who wished to own dogs of these breeds had to apply for a licence to own them (one for every dog).
Applicants would need to give good reasons why they wanted the breed (e.g. “looks” or “I’ve seen a cool dog just like that in a movie” wouldn’t cut it). They’d have to demonstrate that the lifestyle they could offer and where they lived was suitable for the desired breed/breed mix, and they’d have to demonstrate knowledge of the breed and its typical challenges and requirements, as well as general knowledge about dog care and behaviour, and good character… probably through sitting an exam.
The licence should be managed with common sense and flexibility. For example, a new owner wouldn’t be able to obtain a licence to have an LGD in an apartment, and would need to come up with good reasons for why they wanted an LGD, but once they already had the dog, if they later moved into an apartment and it worked out OK, then they shouldn’t have to fear that the licence would be suspended just out of principle. It should also be easier for people to obtain licences for new dogs, especially for the same breeds, when they had proven their suitability through the application process once.
There could be a check-up after the first year and then five years, where the owner would be required to demonstrate the dog’s training, and show that the dog was well looked after, mentally sound, and well managed.
Preferably, the “Restricted” category would include all largish, guard-ish and dog-aggressive-ish types of breeds (and their mixes), which might pose a risk to people and other dogs, beyond a few puncture wounds. The definition of “breed” shouldn’t be the kennel club definition, AKA only closed breeds, but a more functional definition that included dynamic, open breeds (landraces and function-bred mixes), such as European Sled Dogs.
It shouldn’t be too hard for people to be approved to own for example a Rottweiler or German Shepherd, if they could explain why they needed a guard dog for protection or dog sports, but it would be extremely difficult for them to convince the assessor that they needed and could manage a Fila Brasileiro or Ovtcharka under Danish conditions, so the framework wouldn’t be more lenient than the current breed ban.
Of course, in principle such a framework could include all breeds, and make it mandatory to obtain a licence in order to own any dog. That would probably improve pet dog welfare and reduce dog-related risks overall, however I don’t think it would be reasonable to require e.g. old people to sit an exam (and perhaps even pay a fee for it) in order to buy a Lhasa Apso, or some other breed that’s extremely unlikely to pose a serious safety concern.
“The effect of breed-specific dog legislation on hospital treated dog bites in Odense, Denmark—A time series intervention study” by Nilson, F., Damsager, J., Lauritsen, J., & Bonander, C. (2018).PLoS one, 13(12), e0208393. https://doi.org/10.1371/journal.pone.0208393
“Breed Specific Legislation Had No Effect on Dog Bites in Odense, Denmark” by Zazie Todd, PhD, on 13 February 2019, in Companion Animal Psychology.
“Evaluering af hundelovens forbudsordning (LOV nr. 717 af 25/06/2010)” by Iben Meyer and Björn Forkman, 12. august 2013. Det Sundhedsvidenskabelige Fakultet, Københavns Universitet.
“Betænkning om farlige hunde“. Betænkning nummer 1514. Delbetænkning afgivet af
Justitsministeriets Udvalg om hunde. Retsudvalget 2009-10.
Hundeloven (Danish BSL of 2010). Miljø- og Fødevareministeriet. Fødevarestyrelsen.
Bekendtgørelse om forbud mod hold af særligt farlige hunde (Danish BSL of 1991, historic) https://onlaw.dk/bekendtg%C3%B8relse/bek-nr-748-af-14-11-1991