Real numbers prompt hard questions

From Animal Cruelty Does Not Predict Who Will Be A School Shooter by Hal Herzog. An article interesting in its own right but also in some of the information it conveys.
le
I love animals and have had pets all my life. I have fostered fledglings and other rescued animals. I suffer great anguish when animals suffer. I regard this as simply a part of the general condition of human empathy. It does not, I don't think, color my view of others who have no interest in animals and pets. Our animal companions are a burden, it is not unreasonable to not take that on. But I would expect most people to have at least some empathetic response to animal suffering though that response might take many forms.

The thought experiment prompted by this article is along the following lines. Would it be reasonable to exclude from your business and/or social life people who indulged in deliberate and intentional cruelty to animals?

My first response would be, Of course! It is analogous to asking, Would it be reasonable to exclude from your business and/or social life a person who was a habitual spouse beater? Of, course you would. All the arguments are readily available at hand to justify the proposition. Danger, absence of empathy, etc.

But think about that answer. You assume no one you know is intentionally cruel to animals (or spouses). You assume, "Everyone I know is good and innocent of that accusation." And social self-selection and affiliative groups do make it more likely that like attracts to like and that your assumption will be true.

How prevalent is intentional cruelty to animals? 1% of the population? 5%? 10%. Surely not more than that.

Herzog reports:

To understand rates of animal abuse among school shooters, we first need to examine the strength of the link between animal cruelty and human-directed violence. Many investigators have compared rates of animal abuse in violent criminals and in people with no history of violence. In 2016, Dr. Emily Patterson-Kane used a statistical technique called
meta-analysis to combine the results of 15 of these studies. She found that 34% of violent offenders had a history of animal abuse. But so did 21% of non-violent individuals in the control groups. Patterson-Kane concluded these differences in abuse rates were, from a statistical point of view, real but small. Indeed, she is more impressed with the fact that most people who commit violent crimes against humans do not have a history of violence directed at animals.

Further, animal cruelty is surprisingly common in “normal” people. For example, studies have found that nearly 30% of college students admit to having committed some form of animal abuse. Indeed, in a recent paper, the psychologists Bill Henry and Cheryl Sanders concluded “Some participation in animal abuse may be considered normative in American males.”

In short, most violent criminals do not have a history of animal abuse, while a large percentage of apparently normal people do.

30% of college students have a history of animal abuse? That makes it reasonably probable that you assumption that no one in your affiliate groups has a history of animal abuse.

It is real easy to automatically conclude that it would it be reasonable to exclude from your business and/or social life all people who indulged in deliberate and intentional cruelty to animals? It's easy because you don't think the rule would be pertinent to your circle of friends.

Herzog's numbers suggest otherwise. Does the knowledge that a not immaterial portion of your circle of friends might have some history of animal abuse change your answer?

I don't know. Visceral anger over animal abuse is a fairly bedrock issue for me but who among my friends might I be startled to lose were I to discover such abuse.

Of course, being human we love black and white rules as long as they are not relevant to ourselves. Show us a cost to those rules (unexpected loss of friends) and we will start coming up with subordinate clauses (Oh, it was only when they were a teenager; Oh, it was accident; Oh, they were drunk, etc.) to carve out exceptions to the rule.

No answer to the rumination but it was an interesting catalyst to examining assumptions.

The article is interesting in itself and worth a read. Lots of, to me, surprising information. For example, from a study of 23 mass school murders between 1988 and 2012.

Consistent with other studies, they reported that most of the shooters (57%) had no history of animal cruelty. They did, however, find that the types of cruelty committed by the perpetrators who were animal abusers were often different than the cruelty committed by “normal” animal abusers. In nine of the ten cases, the animal abuse was “up close and personal.” That is, the acts involved hands-on direct contact with the animal. Indeed, only one of the school shooters had used a method of torment which did not require touching the animals. In seven cases, the abuse was directed at dog and cats. But in no case did the animal cruelty involve the shooters' personal pets or even animals in their neighborhood.

Ok, that seems to be a signal. On the other hand, there is this.

But here is the big surprise. Arluke and Madfis found that four of the school shooters had a record of pronounced empathy and affection for animals. Sandy Hook Elementary School killer Adam Lanza, for instance, apparently became a vegetarian because he did not want to harm animals. And Charles Andrew Williams who killed two classmates and wounded 13 others became extremely upset when one of his friends killed a frog. Whit­­­­­­­­man would keep field mice as pets and even have little funerals for them when they died.

Also a related article, How reliably does animal torture predict a future school shooter? by Arnold Arluke. From the concluding two paragraphs:

Future tipsters and law enforcement workers should be cautious, though, about assuming that animal abuse is a necessary precondition for a shooter. In fact, a systematic investigation of targeted school shootings revealed that there simply “is no accurate or useful ‘profile’ of students who engaged in targeted school violence,” whether demographic, psychological or social. In many cases, animal abuse isn’t, and when it’s there, it takes a specific form. And even with research-based checklist items such as animal abuse, some cases will inevitably not be red-flagged. If we flag every incident of animal abuse, our mental health and justice systems would be overwhelmed with tens of thousands of cases to review.

It is understandable that people will call for more reliable checklist items that can spot future killers and prevent school massacres. As Adam Lankford, a criminologist at the University of Alabama, lamented: “In many recent cases, law enforcement has known about a future mass shooter before his attack but failed to recognize the threat he posed. With more accurate warnings signs, they may be able to prevent the next tragedy.” Given the complexity of predicting any human behavior, let alone extreme killing, more precise warning signs will not identify every future shooter. But our research offers hope for spotting warning signs of — and thus preventing — at least some school shootings.

We would wish to reduce human violence in all forms to near zero. But human systems are complex and the fact that extreme forms (such as school massacres) are so rare, it means that it is near impossible to create a rigorous algorithm that narrowly and accurately forecasts those likely to commit such acts.

As soon as we relax the parameters even a little, the pool of candidates explode logarithmically and overwhelm the relevant systems (security, mental health, school, etc.). It also overwhelms of foundational principles of governance such as freedom, privacy, negative rights, rule of law, etc. Big data systems and data collection around the world (so that we have a more statistically reliable pool of data) might help us in the future but we seem a long way from a state of cognitive grace at the present time.