The Fallout of Visual Breed Identification

October 4, 2024

Visual identification - determining a dog's breed makeup based on their appearance - is commonly used in both animal welfare and law enforcement, despite ample evidence that it is highly inaccurate. Inaccurate breed labels can have a negative impact on dogs in animal shelter environments and beyond, with little to no benefit to dogs or humans.

What is visual identification?

Visual identification is the practice of labeling a dog as a particular breed or breed mix based on their appearance. Relevant characteristics may include the dog's size, weight, stance, body conformation, musculature, legs, paws, chest, tail, coat length, coat texture, coat color, head/skull shape, face, and/or ears. [2, 6]

In a study published in 2014, participants reported that they may assign a breed label through visual identification based on their own previous experience with dogs, breed standards for a particular breed, and/or similarity of the dog's appearance to another dog who has been labeled by another shelter staff member. [2]

When and where is visual identification used?

Visual identification is commonly used by staff and volunteers in animal shelters and rescues; veterinary practices, including veterinarians and vet techs; law enforcement, including police officers and animal control officers; and by individual pet owners.

Many animal shelters and veterinary practices use management software that requires staff to label each dog as at least one breed, sometimes without the option to simply label a dog as a "mix." Dogs with no documented history, such as dogs who come in as strays, or who have no documented ancestry are typically given breed labels based on visual identification. These labels will often stay with the dog for the remainder of their life.

When it comes to dog bites and Breed Specific Legislation (BSL), visual identification is often used by law enforcement, including both police and animal control officers. In some cases, a dog bite victim or a bystander may use visual identification to report the dog's breed. These breed labels may determine whether a dog is allowed to live in a particular community, must conform to specific legislation (must wear a muzzle in public, must be sterilized, etc.), or is euthanized.

The inaccuracy of visual identification

In general, visual identification is highly inaccurate, especially when it comes to determining two or more breeds in a mixed breed dog with no documented ancestry. [1, 2, 4, 6, 7]

With the rise in popularity and accessibility of genetic breed tests, the inaccuracy of visual identification has only become more apparent. Most genetic tests trace a dog's lineage back to their great-grandparents, listing each detectable breed by percentage. See an example of genetic test results below.

DNA test results for a dog named Santi, including 40.3% American Bully, 24.6% American Pit Bull Terrier, 12.3% German Shepherd, 8.4% Rottweiler, 79% Supermutt, and 6.5% Bulldog

In a study published in 2018, 919 dogs between two US shelters were administered genetic breed tests. Staff members were able to accurately label a dog's primary breed about 56% of the time, including purebred dogs, but identified both the primary and secondary breed only 10.4% of the time. Only 44.5% of dogs had a breed makeup in which a single breed made up 50% or more of the dog's ancestry, with a total of 18.7% of dogs identified by a single breed, two specific breeds, or one breed plus “mixed.” The remaining 81.3% had a breed makeup that included 3 or more distinct breeds. 33% of dogs labelled as pit bull mixes had zero pit-type breeds detected. [6]

A study published in 2015 found that veterinarians were no more likely to accurately label a dog's breed makeup than any other type of shelter staff. [1] Similarly, visual identification in dog bite reports is so inaccurate that the Center for Disease Control (CDC) stopped collecting data on dog breeds related to bite incidents in the 1990s. [8] This suggests that the inaccuracy of visual identification is inherent, not an error of the person conducting it.

How much does breed affect behavior?

When it comes to purebred dogs, evidence suggests that the typical behavior of specific breeds has changed over the past several decades as dogs have increasingly been bred for companionship than for work or show, often with a focus on appearance. Because behavior is influenced not only by genetics but by environment, socialization, and a host of other factors, dogs of the same breed present a wide variety of behaviors. In mixed breed dogs, a dog's breed makeup is not an accurate predictor of behavior. [3,5]

Breeding for appearance has also "inadvertently selected for specific genetic diseases" in purebred dogs. [3] These genetic diseases are less likely to be found in mixed breed dogs, even if the dog has a primary breed that makes up 50% or more of the dog's breed makeup.

With this information, it's reasonable to conclude that a dog's breed makeup has little impact on the dog's overall behavior, especially when it comes to mixed breed dogs. It is much more productive to consider the behavior of the individual dog than it is to make assumptions based on appearance or documented breed.

Scott and Fuller: Paving the way

In the 1960s, 'Genetics and the Social Behavior of the Dog' detailed a study in which several generations of Cocker Spaniel/Basenji mixes were evaluated for learned behavior, inherent behavior, and appearance. As one of the first studies of its kind, it is now considered a classic, often used as a reference in modern studies on breed and behavior.

By the second generation, the mixed breed dogs displayed a large variety of colors and physical attributes, with none of the 72 offspring closely resembling either Cocker Spaniels or Basenjis (see a few of the dogs in the images below, taken from Genetics and the Social Behavior of the Dog'). The offspring would be as likely to be identified as "Labrador Retriever, Beagle, or Springer Spaniel mixes" as their parental breeds. [3] By today's standards, some of the dogs may have even been labelled as pit bull mixes.

The study concluded that the parental breeds were not an accurate predictor of the offspring's behavior, with factors such as environment and socialization playing a role in each dog's behavior. In the following decades, many more studies have come to the same conclusion: behavior is complex and cannot be reduced to a single contributing factor

Six images of puppies standing and looking to the right, with straps holding them in position. Each dog looks unique from the rest.

A black and white image of two dogs, both of which appear to be black with white markings, look at the camera.

Visual identification and pit bulls

The term 'pit bull' is commonly used to refer to as many as a dozen distinct breeds, including the American Pit Bull Terrier, American Staffordshire Terrier, Staffordshire Bull Terrier, American Bully, and others, as well as mixed breed dogs who may or may not have a distinct 'pit-type' breed in their breed makeup.

In animal shelters and rescues, dogs with an unknown history are frequently labelled as pit bull mixes based on their appearance - often a wide head, short fur, and/or a stocky build. When those same dogs are administered a genetic breed test, as many as 1/3 of the dogs labelled as pit bull mixes had zero pit-type breeds in their genetic makeup as far back as their great-grandparents. [1, 2, 6]

In a paper published in 2016, a series of four studies were conducted to evaluate how labelling shelter dogs as pit bulls impacted the way that potential adopters viewed them. When dogs were labelled as pit bulls, they were deemed less attractive, were less likely to be adopted, and had a longer length of stay in the shelter. When breed labels were completely removed, those same dogs were perceived as more attractive to adopters, increasing the likelihood of adoption and reducing the amount of time that the dogs stayed in the shelter. Pit-type dogs weren't the only dogs who benefitted from removing breed labels, either; all breed groups saw an increase in adoption, especially working and herding groups. [7]

Breed labels impact dogs outside of the shelter, too. Though ineffective at reducing the number of dog bites and dangerous dog incidents, Breed Specific Legislation (BSL), housing restrictions, and even insurance providers target pit-type dogs. [9, 10] These restrictions impact the ability of families to find affordable housing with a dog who has been labelled as a pit bull, which both discourages adopters from bringing home a shelter dog labelled as a pit bull and increases the likelihood that a family will re-home or surrender their dog to a shelter.

Alternatives to breed labels in shelters

Given evidence that visual identification is highly inaccurate and that breed labels can negatively affect shelter dogs, removing breed labels entirely and shifting the focus to the individual dog's personality and behavior traits could be beneficial for the dogs, the shelter, and adopters.

Shelters may describe a dog's comfort around other animals or people, activity level, preferred activities, and/or other personality or behavior traits to help adopters determine whether a dog is the right fit for them. Utilizing volunteers and short-term foster homes who can spend one-on-one time with each dog may be one of the best ways to gather this information. Shelters should also make it clear to adopters that every dog will need time to decompress in their new home, and that changes in behavior are common as the dog settles in. Providing print, online, and training resources to new adopters can help improve owner retention.

In recent years, alternatives to the term 'mixed breed' have been suggested for dogs with unknown backgrounds or undocumented lineage, including 'all-American dog' and 'American shelter dog.' [3,5] Other regional terms are used by individual organizations, such as 'Detroit special' or 'MexiMutt.' While these types of terms may serve to bridge the gap for adopters who are curious about a dog's breed makeup, they may not support efforts to focus on the individual dog's personality and behavior as well as removing breed labels altogether.

Conclusion

With decades of research displaying the inaccuracy of visual breed identification, the impact of mislabelling dogs, and the lack of predictability in behavior based on breed, it's time to shift the way that we label dogs, especially in a shelter environment. Putting the focus on the individual dog can increase adoptability and successful adoptions and decrease discrimination through BSL and housing restrictions.