But the conclusions the funders of the study are drawing from it are not only wrong, they’re pernicious.
A new study, by researchers at Mississippi State University, claims the number of dogs killed in U.S. shelters is down to 776,970 (with an upper range of 866,377). If true, this is incredibly good news. The latest data available from the No Kill Advocacy Center’s review had put the number at about 1.1 million, 22% of all dog intakes. It would represent the first time in the modern era that dog deaths dropped below 1 million.
But there are several reasons to view this study with a grain of salt. This was an industry funded study by groups such as Petland, the American Kennel Club, and the Pet Industry’s lobbying group, PIJAC. These groups aim to thwart restrictions on their industry, including the growing trend of banning the retail sale of commercially bred puppies. They are apparently using the study to thwart those efforts and to get policy makers to embrace commercial breeding by claiming that there are not enough dogs to meet demand. The old admonition to “follow the money” would therefore warrant close scrutiny of the study’s methodology and its subsequent results. According to the Center for Accountability in Science, industries “can design research that makes their products look better. They can select like-minded academics to perform the work. And they can run the statistics in ways that make their [conclusions] look better than they are.”
Funding bias, according to Wikipedia, “refers to the tendency of a scientific study to support the interests of the study’s financial sponsor. This phenomenon is recognized sufficiently that researchers undertake studies to examine bias in past published studies.” It has led to tainted studies involving sugary soft drinks, tobacco, and drugs. It can apply to the commercial breeding industry.
Admittedly, that alone is not reason to discard it as industry funding does not, in and of itself, mean there is funding bias or that the study’s conclusions are erroneous. Plus, I want it to be true. It would be good if it was true. If it was true, it would mean more than 300,000 fewer dogs are being killed that previously thought. And that is news we should all celebrate.
But, in fact, there are problems with how the study was conducted. First, the authors use the “capture-recapture” method (used for counting wildlife) in a novel context (shelter killing) and for it to be accurate, the number of shelters they surveyed have to be truly representative of the whole. That means they have to be evenly distributed in terms of geography, size, services, and population. They do not seem to be. For one, the authors only heard from 14% of the shelters they surveyed from 49 states. Put another way, 86% of the shelters did not respond, nearly nine out of 10. Moreover, it is not clear that the responding shelters were representative given self-selection bias: the responding shelters may be those who keep good data, are generally responsive and therefore atypical, and/or are proud to share their data. Our most poorly performing pounds tend to be poorly managed, unwilling to keep good records and averse to transparency. In fact, laws forcing shelters to make their statistics public are often required and when introduced, meet fierce opposition, suggesting that those killing large numbers of dogs would be the least likely to share their data with outsiders if they do not have to.
The study also does not indicate which state they excluded, but a large state could add hundreds of thousands to the total. In addition, a state like Delaware, with a statewide live release rate of 91% for dogs but which does not accept family-relinquished dogs, would skew it down, given the small sample size.
Study authors also excluded foster-based shelters and rescue groups and while that is fine to determine how many dogs are killed in brick and mortar shelters, it is not fine with determining the number of dogs available and therefore the conclusions they want to draw from the study.
For example, the study authors started with 10,890 shelters and rescue groups. Of these, they removed 114 duplicate listings and 40 additional agencies which did not shelter dogs. That left them a sample size of 10,736 organizations. They then also excluded 7,649 rescue groups and foster-based agencies. And they also excluded shelters which refused to do adoptions (which means greater killing, meaning they took out those shelters that would push the numbers killed highest).
Of the 10,736 organizations (excluding duplicates and those who did not shelter dogs), they drew up a list of 3,064 shelters to review, 29% of the total. Of those, they reduced it even further to 2,862 (from 49 of 50 states) based on various problems. And of those, only 413 responded, 14% of their sample size, but only 4% of the overall total of organizations.
By contrast, the No Kill Advocacy Center looked at reporting data from all mandatory reporting states including, but not limited to, California, Michigan, Virginia, Colorado, and North Carolina, a database of over 1,000 shelters, and the databases of other organizations captured via self-publishing and Public Records Act results to get 1.1 million. But assuming for the sake of argument that the industry funded study is the more accurate of the two—and because it means over 300,000 additional dogs are spared, I want theirs to be true—it still wouldn’t lead to the Pet Industry’s conclusions about it. In fact, you could take the exact same study and the exact same results and conclude the opposite: that with 776,970 dogs still being killed in shelters, creating more dogs that would compete for homes with those facing death is immoral. I certainly would.
The industry group which funded the study has side-stepped this conclusion by claiming that the vast majority, if not all, of the dogs killed in shelters are non-rehabilitatable. They are not. Looking at the results of the best performing shelters in the nation and pioneering research in behavior and shelter medicine, 99% of dogs in shelters are not irremediably suffering. Of the 5,532,904 dogs who entered shelters, only 55,329 dogs would meet the definition for true “euthanasia.” That means that over 720,000 dogs are still losing their lives in shelters but for adequate efforts to find them homes (almost 800,000 at the upper range).
Add to that all the dogs being cared for by rescue groups, dogs in high intake jurisdictions that could be transferred to higher demand jurisdictions, community dogs, dogs available from individuals through neighbors, friends, work colleagues, and newspaper ads, dogs in U.S. territories like Puerto Rico and the U.S. Virgin Islands that need homes, and once we saved all of those, dogs in countries on or near our borders still dying or living on the street, and as I told a Washington Post reporter writing about this issue, we’re nowhere near concluding that breeding would be required to meet public “demand” for dogs.
In short, there is no existing or coming dog shortage. The conclusions drawn from the study are a classic example of a “solution” in search of a problem. To be sure, we do not have to kill dogs as shelter supply is outstripped by demand. But we do not have to breed them either.
Ignoring this, some of the groups which funded the current study have made two additional, disingenuous arguments. First, they argue that people do not want the dogs currently dying in shelters; that they want puppies. Second, they argue that the remaining dogs dying in shelters are dangerous. Neither stands up to scrutiny.
Despite more widespread and affordable sterilization, puppies are in no danger of disappearing from the U.S. And, of course, we can do a better job of marketing the ethics and benefits of adopting young adult, adult, and mature animals to change community preferences. Given the increasing success of the No Kill movement, the cliché that no one will adopt out certain breeds, adults, or animals with impediments has been thoroughly debunked.
While the vast majority of dogs entering shelters are young (the average age is two years old) and healthy, even those who do are finding homes in those cities were the shelter has embraced a culture of lifesaving. Animals are being saved and finding homes, regardless of perceived “breed,” whether they are young or old, healthy or sick, unweaned, injured, or traumatized.
Moreover, a recent study adds to a growing body of literature that should put to rest, once and for all, the false notion that dogs in shelters are in shelters because there is something wrong with them: “Nothing in the prevalence estimates we reviewed suggest that overall, dogs who come to spend time in a shelter (and are not screened out based on history or behavior at intake or shortly thereafter) are dramatically more or less inclined toward problematic warning or biting behavior than are pet dogs in general.”
Given that far less than 1% of pet dogs bite people, the conclusion is inescapable: shelter dogs are not dangerous. In fact, looking at bite rates that require hospitalization, only 0.001% of dogs (or roughly 1 in 10,000) bite with enough force to cause an injury.
These studies mirror the findings of the most progressive and successful municipal shelters (and those running “open admission” shelters under contract) in the country: those saving 99% of dogs. Doug Rae, the director of the Humane Society of Fremont County in Colorado, is instructive. The Fremont Humane Society runs the animal control shelter for seven cities under contract. Rae also previously ran a private shelter in Rhode Island and “open admission”/municipal shelters in Arizona, Indiana, Maryland, and Pennsylvania, including those taking in as many as 30,000 animals per year. Says Rae:
Over the years, I have rarely seen a truly ‘aggressive’ dog. The vast majority are simply scared. I was speaking with a Board member for a shelter on the West Coast last week and she said that her shelter saves 96-97% of the animals and that 3 or 4% are dogs who are aggressive and need to be killed. In my experience, the percentage of truly aggressive dogs I have seen in small to very large shelters is well under one quarter of 1%.
Ironically, research into dogs from Commercial Breeding Establishments (CBE) shows the opposite; that these dogs show deep psychological scarring as a result of the trauma they experience at the facilities. Compared to shelter dogs, commercially bred dogs exhibited more fear, nervousness, health problems, compulsive behaviors, house soiling, and sensitivity to touch. In some cases, significantly more. Many of these dogs experience “regular and often persistent fear or anxiety, even after years in their adoptive households” as a result of stress-induced psychopathology and inadequate socialization. These dogs have been psychologically damaged. And their offspring may also suffer:
Offspring of pregnant animals exposed to [these kinds of..] stressors have been documented with neurohormonal dysfunction… impaired ability to cope with stress; exaggerated distress responses to adverse events; impaired learning; abnormal social behaviour; increased emotionality and fear-related behaviour; and fearful behaviours that increase with increasing age; increased susceptibility to pathophysiological outcomes when further adversity occurs during adulthood; and behavioural deficits and molecular changes in the offspring similar to those in schizophrenic humans. (Citations omitted.)
A subsequent review of the literature compared the behavior of dogs obtained from pet stores and/or born in breeding establishments and compared them with dogs from other sources to determination causes of behavior problems that occur disproportionately in pet store/CBE dogs. A review of studies over the last two decades finds, “that dogs sold through pet stores and/or born in high-volume, commercial breeding establishments (CBE) show an increased number of problem behaviors as adults.”
The findings included:
- Aggression to people was more than twice as likely in dogs acquired from pet stores compared to those acquired from shelters;
- Dogs acquired from pet stores were more likely to develop social fears (of strangers, children, and other dogs) than from all other sources
- Dogs acquired from pet stores were more likely to be separated from their mothers at a young age and these dogs had a four-fold increase in destructive behaviors
- Dogs acquired from pet stores were more excitable, less trainable, had increased separation-related behaviors, escape behavior, and sensitivity to being touched;
- Dogs acquired from pet stores were more likely to house-soil.
Why? There were several potential causes: Genetics, prenatal developmental stressors impacted by the experiences of the mother in the puppy mill, stressors as a result of early life in the CBE, early maternal separation, transport stresses, lack of adequate socialization early in life, pet store experiences, and the number of relocations early in life. The study was not able to pinpoint specific causes to individual dogs because all the pet store dogs were subjected to “high number of stress-related factors” which impact behavioral development.
In layman’s terms, CBEs engage in systematic neglect and abuse toward animals, causing severe emotional scars on the victims. The former breeding dogs living with those scars are a testament to that fact. Not only do one in four have significant health problems, many of them are psychologically and emotionally shut down, compulsively staring at nothing.
Even if we accept the industry’s argument that there is a difference between “puppy mills” and “responsible commercial breeders,” dogs and puppies are not commodities. At the very least, they shouldn’t be. Not only because the trade in sentient beings is unethical, but because there’s a supply chain that can never fully be. When there is a profit to be made on the backs of dogs, those backs are strained and often broken.
The Washington Post article is available by clicking here.
If you would like a copy of “the study,” please contact the No Kill Advocacy Center.
For further reading:
Have a comment? Join the discussion by clicking here.