No Animal Left Behind
What the science says about running a humane shelter
This is an updated version of an article that was published in 2022. The current article has several new studies, expanded discussion, alternatives to temperament testing, links to an updated matrix of treatable behavioral and medical conditions, a discussion of rabbits and other non-dog/cat species, and additional recommendations.
In the dog room
Shelters are stressful places for dogs and can be stressful places for potential adopters. In most shelters, dogs can’t see people or other dogs in neighboring kennels. Where there is glass, it’s often opaque. Where there are fences or bars, they face a wall. When people visit, they are told not to touch the animals. These are mistakes that increase frustration for dogs.
When frustrated, dogs are stressed, bark excessively, and engage in antisocial behavior. The louder the barking, the less time people spend in the kennels looking for dogs to adopt. And the more dogs bark or act out, the longer it will take them to be adopted.
Moreover, studies have found no compelling evidence that most dogs are in shelters because of something wrong with them. Instead, “surrenders often say more about the people doing the surrendering — about ‘owner-related factors, needs, and expectations’ — than the dogs being surrendered.” As such, shelters should stop thinking of dogs as having “behavior problems” and instead understand them as “behavior incompatibilities” with the person they were living with before being surrendered.
For example, like human children, dogs go through a rebellious adolescent phase. Although enrichment/reward-based training can modify these behaviors, research has found that “this corresponds with the peak age at which dogs are relinquished to shelters.” Not surprisingly, most shelter dogs are young. Without enrichment in the shelter, they may be falsely labeled “aggressive” or “unadoptable” and killed.
Various studies conclude that “Even in well-managed and funded facilities, dogs are likely to encounter an array of stressors including noise, unpredictability, loss of control… disruption of routines…” and unfamiliar people and surroundings. Many dogs fail their behavior evaluation without rigorous efforts to remedy these issues. By contrast, these dogs will likely pass if provided enrichment, including those initially, though falsely, deemed “potentially quite dangerous” by shelter staff.
These enrichment activities include allowing dogs out of the cage/kennel for human interaction, dog-dog play, several daily walks, letting dogs see people and dogs in shelters by removing visual barriers that block such access, and group housing. Over one-third of dogs housed alone suffer behavior problems, and 10% engage in repetitive behavior (like endless barking). By contrast, dogs housed with other dogs have “a decrease of stress after long-term stay in the shelter.” As such, dogs in shelters should be pair or group-housed if they get along with other dogs. Aggression and fighting are rare if the staff is thoughtful about pairing.
Finally, conventional wisdom says the longer dogs are in the shelter, the more likely they are to become “kennel crazy” and thus “less adoptable.” This is false as “dogs adapt to the kennel environment over time” and “environmental enrichment helps animals to cope with their environments.” In other words, newly admitted dogs tend to be stressed. Dogs who only get the basics — food, water, and shelter — are stressed. However, socialized and exercised dogs are not stressed; the longer they are in the shelter, the less stressed they become.
Findings: Requiring an appointment to adopt animals leads to significantly fewer adoptions — and fewer adoptions lead to more killing.
Study: “What is the Evidence for Reliability and Validity of Behavior Evaluations for Shelter Dogs?”
Findings: There is “no evidence that any canine behavior evaluation has come close to meeting accepted standards for reliability and validity.” Some tests are wrong as much as 84% of the time (a combination of poor tests and poor testing practices by pound workers). While shocking, it should not be surprising since the tests rest on a “fatally flawed” premise: “that the provocations used at a single time during a dog’s stressful experience in a shelter will predict future behavior at a different time and place.”
Study: “The Relationship Between Aggression and Physical Disease in Dogs.”
Findings: Medical conditions “can trigger aggression in a nonaggressive dog.” This includes hyperthyroidism, neurologic issues, exposure to distemper, skin problems, and microbiome changes. Specifically, “aggression can be a defensive reaction to avoid a particular handling that the dog anticipates as potentially painful.” Moreover, “the stress response associated with pain reduces serotonin activity,” and a “reduction in serotonin activity in the [central nervous system] has been linked to aggressive behavior in dogs.” Animal shelters that claim a dog is “aggressive” must “include a thorough physical examination to rule out painful conditions, a neurologic examination, a complete blood count, and biochemistry.”
Findings: Shelter staff cannot accurately predict dog behavior in a home based on dog behavior in a shelter: “displaying concerning or dangerous behavior was not significantly associated with return to the shelter, and positive predictive values were low, implying that many positive tests will be false positives.” Indeed, “dangerous behaviors” occurred in sheltered dogs significantly “less than 5%” of the time.
Findings: Shelters are very stressful places for dogs, causing them to fail behavior evaluations. “Even in well-managed and funded facilities, dogs are likely to encounter an array of stressors including noise, unpredictability, loss of control… disruption of routines…” and unfamiliar people and surroundings. A small amount of enrichment — being spoken to softly, given treats, petted, and played with — can result in dogs passing temperament tests. After just five days of being treated kindly, “nearly all” fearful dogs passed the test. This is true even for dogs deemed “potentially quite dangerous” at the beginning of the study.
Study: “Characteristics and Adoption Success of Shelter Dogs Assessed as Resource Guarders.”
Findings: While 15% of kennelled dogs guard their food, many “do not guard food in their adoptive homes, and, even when dogs continue to display food guarding in the home, adopters do not consider it to be a major problem.” As such, shelters should opt for “adoption rather than euthanasia for most dogs identified as resource guarders during behavioral evaluations in shelters.” Dogs returned for food guarding should be readopted and can be adopted without incident.
Study: “Saving Normal: A new look at behavioral incompatibilities and dog relinquishment to shelters.”
Findings: There is no compelling evidence “for the notion that the general population of relinquished dogs in shelters are there because of relationship-breaking behavioral incompatibilities in their prior home.” Most dogs labeled “behavior” are normal as “surrenders often say more about the people doing the surrendering — about ‘owner-related factors, needs, and expectations’ — than the dogs being surrendered.” As such, shelters should stop thinking of dogs as having “behavior problems” and instead refer to them as “behavior incompatibilities” with the person they were living with before being surrendered.
Findings: “[T]he welfare consequences of adolescence-phase behavior could be lasting because this corresponds with the peak age at which dogs are relinquished to shelters.” In other words, like human children, dogs go through a rebellious adolescent phase, and reward-based training can modify these behaviors.
Study: “Impact of Visual Barrier Removal on the Behavior of Shelter-Housed Dogs.”
Findings: Letting dogs see people and dogs in shelters by removing visual barriers that block such access reduces stress in shelter dogs. Stressed dogs are likely to fail their behavior evaluations.
Study: “An Evaluation of Respondent Conditioning Procedures to Decrease Barking in an Animal Shelter.”
Findings: Poor behavior in shelter dogs is often a sign of frustration. Specifically, dogs can see people and other dogs but can’t interact with them. Over time, they associate people and other dogs with frustration. Dog treats counter-conditioned them. It doesn’t matter if dogs are given treats when they weren’t barking, were barking, or when conditioned to identify a sound — such as a door chime — all three improved dog behavior.
Findings: Group housing dogs reduces frustration. Over one-third of dogs housed alone suffered behavior problems, and 10% engaged in repetitive behavior (like endless barking). As such, dogs in shelters should be pair or group-housed if they get along with other dogs. Fears about aggression and fighting in pair-housed kennels tend to be overblown, with fights being rare so long as the staff is thoughtful about pairing.
Findings: Dogs co-housed with other well-matched dogs are less stressed and adopted faster than those housed alone.
Findings: Conventional wisdom says the longer dogs are in the shelter, the more likely they are to become “kennel crazy” and thus “less adoptable.” This is false as “dogs adapt to the kennel environment over time” and “environmental enrichment helps animals to cope with their environments.” In other words, newly admitted dogs tend to be stressed. Dogs who only get the basics — food, water, and shelter — are stressed. However, socialized and exercised dogs are not stressed; the longer they are in the shelter, the less stressed they become.
Study: “Welfare Assessment in Shelter Dogs by Using Physiological and Immunological Parameters.”
Findings: “During the initial physiological adaptive response, novel environment, manipulation, change in social structure, and different cages can represent sources of stress for dogs entering a shelter,” but for dogs housed with other dogs, the result is “a decrease of stress after long term stay in the shelter.” In other words, dogs should be able to see and interact with other dogs and people to “reduce frustrated attempts to see what is going on beyond their kennel.”
Study: “Ancestry-inclusive dog genomics challenges popular breed stereotypes.”
Findings: Breed tells us how dogs look, not how they behave. For “predicting some dog behaviors, breed is essentially useless, and for most, not very good.” For example, “the defining criteria of a golden retriever are its physical characteristics — the shape of its ears, the color and quality of its fur, its size — not whether it is friendly.” Likewise, the findings “cast doubt on breed stereotypes of aggressive dogs, like pit bulls.” The study adds to a growing scientific literature that demonstrates breed bans are useless for public safety and breed descriptions are useless for matching dogs who exhibit desired behaviors with prospective adopters.
Study: “For Pits’ Sake: Effect of Breed Labels on Perceptions of Shelter Dog Attractiveness.”
Findings: Shelter workers often misidentify breeds: 50% of dogs labeled “pit bulls” lacked DNA breed signatures of breeds commonly classified as pit bulls. Moreover, perception of breed negatively impacts length of stay and adoption rates. Consequently, removing these labels allows these dogs to be adopted and adopted more quickly. It also results in a better fit between the family lifestyle and the individual dog’s behavior.
Study: “Perceived importance of specific kennel management practices for the provision of canine welfare.”
Findings: Compared to volunteers, rescuers, and the average person, shelter employees were less likely to view health, hygiene, and enrichment as necessary or essential. These false beliefs put dogs at risk: “What people believe is important will influence their behavior, with direct relation to care provided to animals.” Given inconsistent, unenforced, and, in many cases, non-existent regulations that mandate a commitment to dog welfare, what dogs need “may not be successfully translating into evidence-based changes in industry practice.” In other words, unless forced, shelters will not likely ensure dogs are kept healthy, clean, and socialized. Failure to force them to do so puts dogs at risk of being killed.
Study: “A Review on the Influence of Noise on the Welfare of Dogs.”
Findings: Almost half of all dogs fear noise to varying degrees, including the sounds of an animal shelter, and for those “shelters” that rely on arcane “temperament tests,” dogs could be labeled “unadoptable” and killed as a result. All shelters must employ noise abatement strategies “because the welfare of animals may decline” if they do not. This includes “incorporating sound control in the design of… buildings” such as acoustic clouds, using “quieter cleaning tools,” “replacing noisy overhead fans,” and enrichment strategies, including music with soft, human voices. In cases of extreme fear, medication may also help.
Study: “The effect of different genres of music on the stress level of kenneled dogs.”
Findings: Soft music with human voices reduces stress in kenneled dogs. Lowered cortisol levels, heart rate, and stereotypy behavior measured this. Dogs preferred, in order, soft rock, reggae, pop, and then classical. The soft human voice is the most essential part of the musical experience for dogs. However, dogs, like humans, get bored of the same playlist. By the end of the first day, the impact of stress reduction from playing the same songs was eliminated. By day seven, the effect of the same type of music (e.g., classical or soft rock) was eliminated.
Findings: Dogs thrive around other animals: “like many social animals — including humans, having more social companions can be important for the dog’s health.” Isolating dogs leads to poor health outcomes.
Findings: Dogs can smell when humans are stressed, which impacts their choices, resulting in dogs making more pessimistic judgments if exposed to a stressed person.
Note: In addition to these findings, numerous studies have found that the ability of people, including shelter workers, to identify canine behavior correctly is poor, despite most rating their understanding of dog behavior highly. For example, “Barking and growling are not isomorphic [synonymous] with ‘aggression.’” Neither are mouthiness and other behaviors typically labeled as “resource guarding” behavior, “jealousy,” or “aggression.” Often, the dog is just seeking information: “When dogs ask questions of and gain information… in active interactions, tactile exchanges are frequent and complex, often involving mouthing, an activity that humans both misunderstand and actively discourage.”
Not surprisingly, “over one-third (36.9%) of dogs originally brought in by their owners for euthanasia could, upon further evaluation by staff and discussion with owners, be made available for adoption” given that they “had medical or behavioral concerns that were amenable to resolution, as opposed to all having terminal health conditions or intractable behavior problems.”
In the cat room
Shelters are stressful places for cats. Shelters that do not have a “mental health” component (touch, talk, play) in concert with a “physical health” component (vaccination on intake, prompt and necessary veterinary care, cleaning/disinfection) undermine the well-being of cats and put them at risk of getting sick and being killed.
Moreover, many cats fear noise to varying degrees. Barking dogs and other loud noises often trigger this fear in shelters. At shelters that have not embraced a culture of lifesaving, this can lead to cats inaccurately being labeled “fearful,” “fractious,” or “unadoptable” and killed.
All shelters must employ noise abatement strategies to avoid negatively impacting animal welfare. This includes sound control, such as acoustic clouds, quieter cleaning tools, and enrichment strategies. For example, unless housed in a large, free-roaming cat room, cats should be allowed out of the cage/kennel for human interaction, playing, and running at least twice daily. Moreover, cats should be gently petted and talked to daily. These practices result in a markedly lower chance of upper respiratory infection. Cats who are not stroked and talked to gently are over two times more likely to get sick (due to stress) than cats who are. Ironically, many shelters do not allow people to touch cats due to fear of disease, placing signs throughout the shelter to that effect.
Unfortunately, shelter employees are less likely to view health, hygiene, and enrichment as necessary or essential than volunteers, rescuers, and the average person. These false beliefs put animals at risk. Given inconsistent, unenforced, and, in many cases, non-existent regulations that mandate a commitment to animal welfare, studies have found that what animals need “may not be successfully translating into evidence-based changes in industry practice.” In other words, unless forced, shelters will often not ensure animals are kept healthy, clean, and socialized. Giving shelter staff and leadership the benefit of the doubt puts animals at risk of being killed.
Finally, while shelters label roughly two out of five cats “feral” or “aggressive” (and these cats are killed unless the shelter embraces a community cat program), very few cats will respond fearfully after day six if gently petted and talked to throughout their stay in the shelter. This is also true of cats who cannot be touched when they arrive. As such, “a 3-4 day holding period“ is not “sufficient to differentiate non-feral from feral cats.”
Regardless, No Kill shelters do not kill cats for “behavior,” “aggression,” or being considered “feral.” Community cats who are not social with humans are returned to their habitats. Cats deemed “fearful” or “fractious” but otherwise social with humans are adopted out, despite those issues, because cats do not pose physical risks to public safety. The behavior also often resolves outside of a shelter.
Findings: Cats gently petted and talked to have a markedly lower chance of getting an upper respiratory infection. Cats not stroked and talked to gently were over two times more likely to get sick (due to stress) than cats who were. Ironically, many shelters do not allow people to touch cats due to fear of disease, placing signs throughout the shelter to that effect, even though the absence of touching makes them 2.4 times more likely to get sick.
In addition, cats who are labeled “feral,” “unsocial,” “fractious,” or “aggressive” are virtually all killed unless the shelter embraces a community cat sterilization program. While 18% of the cats tested would have been deemed “feral” due to “aggression” (and thus killed), none of the cats responded that way after day six. This is also true of cats who could not be touched when they arrived and were stroked “mechanically” with a fake hand. As such, “a 3-4 day holding period” is not “sufficient to differentiate non-feral from feral cats.”
Study: “The Impact of Lethal, Enforcement-Centred Cat Management on Human Wellbeing.”
Findings: Besides harm to cats, killing “feral”/community cats results in profound human suffering. Cat caregivers are motivated by “considerable concern regarding the health and safety of the cats,” “feel responsible for improving their health and welfare,” and “commit substantial time and finances to their needs, despite existing legal and financial difficulties.” As such, the bond between caregivers and cats was “as strong as the bonds with their own pets,” and “the cats looked to them (the caregivers) to keep them safe and fed.” So, it is not surprising that caregivers suffer when officials kill cats. That suffering was “significant,” leading to grief, trauma, poor physical health, and long-term psychological distress, including profound guilt, loss, and inability to eat. Given this and other evidence, shelters must abandon lethal methods and instead “identify and assist caregivers to implement neutering and, if possible, adoption. This would improve cat welfare, minimize public complaints, and reduce psychological hazards to caregivers.”
Note: Killing a cat for “behavior,” “aggression,” or being considered “feral” should never occur. There isn’t even a need to delay finding homes for these cats. They can be sterilized and returned to their habitats if they are not social with humans and are used to living outdoors, or they can be adopted out immediately if they are. Simply put, people will adopt cats with “cattitude.”
This is not to say that cats who experience behavior issues in the shelter do not warrant changes in shelter housing, shelter treatment, and behavior intervention to address those needs. They do. However, they can be adopted despite those issues because resolving behavior challenges is almost always done by getting them out of the shelter. Moreover, for those who do need further treatment, treatment in the home will be more effective and focused, as it is for many dogs.
As the director of an open-admission animal control shelter, I eliminated any “behavior category” for cats and thus any killing of cats for “behavior,” “aggression,” or being “feral.”
In the rooms with rabbits and other animals
Although rabbits make up the third most common animal in shelters, they — and other non-dog and cat species — are often treated like second-class residents. While all U.S. states have holding periods for stray dogs and cats, they may have none for bunnies, allowing shelters to kill them immediately with no attempt to rehome them.
Moreover, while shelters might have rules prohibiting the adoption of dogs who will not be allowed in the house because lack of socialization and affection are traumatic for dogs, they have no such rules for rabbits who suffer similar harm and likewise belong in homes, not backyard cages. Finally, many shelters do not vaccinate or sterilize rabbits despite doing so for dogs and cats.
The No Kill movement has also primarily ignored bunnies, rodents, “exotic” pets, birds, wildlife, and farmed animals. The need to save rabbits is conspicuously absent on the platforms of many organizations. They have resisted calls to add them to statistics when calculating overall community placement rates. Many shelters placing over 95% of dogs and cats either have no safety net for rabbits and other animals — turning them away — or they kill them in large numbers but still call themselves No Kill.
Even some rabbit advocates do them a disservice by choosing death at the pound rather than have them adopted out to homes with children (buying into the same misanthropy that dog and cat rescuers rejected a decade ago). Adding insult to injury, rabbits are legally considered “food” and abused to test cosmetics.
It is well past time for change.
The No Kill Advocacy Center’s model shelter reform legislation, The Companion Animal Protection Act, gives rabbits and other small animals all the protections of dogs and cats, including holding periods, prompt and necessary veterinary care, and the right of rescue.
Running a humane shelter
A reference manual for running a humane animal shelter. With short encyclopedia-like entries, topics include:
The definition of No Kill;
Costs and benefits of No Kill animal control;
A 2024 snapshot of the No Kill movement;
Sample assessment tools and protocols for running a municipal No Kill shelter;
Animal protection legislation;
Primer on various shelter issues, including temperament testing, community cats, dog behavior, and more.
Based on recent studies, the pioneering work of behaviorists, and the results of some of the most successful agencies in the country, shelters can place 99% of dogs.
Animal Evaluation Matrix: Policies & Procedures That Protect the Lives of Shelter Animals
The Matrix includes medical and behavior protocols, diagnostic tools, and end-of-life protocols, with forms and checklists to increase accountability and improve performance. The No Kill Advocacy Center developed these protocols in collaboration with some of the most successful shelter directors in the country (directors running municipal and animal control-contracted shelters with placement rates of 99%).
Please note: Thanks to advancements in veterinary medicine, The No Kill Companion has an updated list of treatable behavioral and medical conditions.
Traumatized dogs deserve safe harbor and time — time to abandon fear, forget a haunted past, and, most important of all, learn to trust. With the right amount of love, kindness, compassion, positive conditioning, and, when necessary, veterinary intervention, psychologically wounded animals, like humans, have a remarkable capacity for resilience.
A community cat program reduces intake and killing of community cats, illness in the shelter, complaint calls to animal control, spending (and waste of taxpayer money), and increases opportunities to expand lifesaving of other animals, such as dogs.
The No Kill Advocacy Center offers 30 free step-by-step guides for building a humane animal services program.
God Bless you, Nathan!!
Thank you, Nathan, for the latest beneficial changes needed to be made in Shelters -- hope the World takes this to heart.