Animal Abuse and Domestic Violence: Pets as Tools of Control

In a survey that has been cited for decades, 71 percent of women entering domestic violence shelters said their abuser had threatened, injured, or killed their pet. The figure keeps getting quoted not because it is sensational, but because it points at something specific: pets are often instruments of control.

Criminology and social work call this the Link, the observed association between animal cruelty and domestic violence. It is not a causal chain. It is three recurring patterns of co-occurrence, and this piece lays out all three, along with which numbers you can cite and which you should not.

The short answer

In the professional literature, the Link refers to co-occurrence and risk indication, not to the claim that people who abuse animals will abuse partners. It has two practical uses. A threatened pet can serve as an early signal that violence is escalating. And if protection orders, shelters, and law enforcement do not account for pets, victims frequently do not leave.

What the Link actually refers to

Researchers generally break the association into three layers. The first is co-occurrence: within the same household, violence against animals and violence against people often appear together. The second is instrumentalisation: abusers discover that harming or threatening a pet is one of the most effective ways to control a partner, because the partner will compromise in order to protect the animal. The third concerns children: the effects on a child who witnesses animal cruelty overlap with the effects of witnessing violence between people.

None of the three amounts to a causal prediction that animal cruelty leads to domestic violence. The overwhelming majority of people who harm animals never abuse a partner. The value of the Link runs in the other direction. Once domestic violence is already happening, the animal's situation is something observable, something that can be documented, and sometimes the only externally visible signal.

Pets as instruments of control: the mechanism matters more than the numbers

The mechanics of using a pet as an instrument of control recur consistently in shelter interviews, and they take fairly fixed forms.

The most common is a threat rather than an act. If you leave, I will deal with the dog. This works because it turns the thing the victim cares about most into a hostage, and it produces the controlling effect without ever having to be carried out.

Next is the act itself, usually after the victim has tried to leave or called the police, as retaliation or warning. There is also a subtler form: the abuser is kind to the pet in front of the partner and cruel when the partner is absent, which leaves the victim unable to explain what is happening, because what outsiders see is someone who is good with dogs.

The most common outcome, though, is not violence but a structural bind: shelters do not take pets. The victim does not face the question of whether to leave. They face the fact that leaving means handing the dog to the abuser.

The data: what you can cite and what to treat carefully

Secondhand transmission in this area is badly muddled. The same claim gets attached to different years and different studies depending on the source. These are the figures with traceable attribution.

71 percent: women in domestic violence shelters who reported that their abuser threatened, injured, or killed a pet. Source: the National Coalition Against Domestic Violence. This is the most frequently cited figure.

52 percent: survivors in shelters who left their pets with the abuser. Same source. This is where most of the subsequent policy discussion starts.

Nearly half: victims who would rather stay than leave a pet behind. Source: Carlisle-Frank, Frank, and Nielsen, 2004, in a study titled Pets as Pawns.

As many as 25 percent: survivors who return because the abuser is using the pet as leverage. Source: the ASPCA.

One widely repeated claim deserves caution: that only 3 percent of domestic violence shelters accept pets. That number comes from an older survey, and the share of shelters able to take animals has grown substantially in recent years, which is precisely what programs like RedRover's Purple Leash Project exist to push. Cited without a year, it materially understates the present.

The children

Children who witness animal cruelty show effects that overlap with those of witnessing violence between people: trauma responses, damage to basic assumptions about safety, and in some cases the carryover of violent patterns into their own future relationships.

This layer is also where the violence triangle becomes concrete. The triangle refers to three early warning signs: animal cruelty, fire-setting, and bedwetting. The correct use is to trigger attention and assessment, not to label a child. That distinction matters especially in the school context, which we cover in a separate piece.

Policy responses: four mechanisms already running

Over the past decade, policy in this area has become quite specific, and most of it is already in force.

First, pets in protection orders. Most U.S. states now allow courts to include pets in domestic violence protection orders, both by barring the abuser from contacting or taking the animal and by awarding custody of the animal to the victim. This is the lowest-threshold and most directly effective measure.

Second, the PAWS Act. Signed into law as part of the 2018 Farm Bill, it extended federal domestic violence protections to victims' pets and established a grant program covering boarding and emergency placement. It addresses the most practical obstacle of all: wanting to leave and having nowhere to put the dog.

Third, shelters that take pets. Beyond the minority of shelters that house animals directly, the fastest-growing model is cooperative fostering, in which rescue organizations supply foster homes while the shelter handles placement for the person, with grant programs connecting the two. RedRover's Safe Escape Grants and the Purple Leash Project both fall into this category.

Fourth, veterinary identification and reporting. Multiple jurisdictions require or encourage veterinarians to report suspected non-accidental injuries, and some go further, asking vets who see signs of animal cruelty to consider the risk to other members of the household. This one is genuinely contested, because it can deter owners from seeking care, and enforcement varies widely.

If this is where you are: a few concrete things

The most important point first. A pet should not be the reason you stay, and the support system is far larger than it was a decade ago.

Before leaving, assemble the animal's paperwork. Vaccination records, microchip information, veterinary history, photographs. These establish ownership and, if needed, support a claim for custody of the animal. At the same time, prepare a list of emergency contacts, including friends who could take the animal temporarily and local shelters or rescue organizations.

If your local shelter does not take pets, ask four questions directly: is there a partner foster network, is there an emergency boarding grant, is there a pet-friendly shelter in the area, and who coordinates between them. RedRover runs Safe Escape Grants and maintains a shelter map, and some cities keep pet-friendly shelter lists. You do not need to over-explain. These services exist for exactly this situation.

If the animal has already been harmed or taken, preserve evidence and report. Animal cruelty is a separately chargeable offense in many jurisdictions and does not require a domestic violence case to proceed first. We have written about evidence-gathering and reporting separately.

What this has to do with Wangwang

In the Wangwang case, Wangwang was a stray dog with no owner, and the perpetrators were minors who had not reached the age of criminal responsibility. There was no owner whose animal could be claimed as property. That absence is legally fatal. With no animal protection legislation in China, harm to an animal that belongs to no one can only be pursued indirectly, through property loss, and that route is close to useless for an animal with no owner.

That is exactly where the Link matters. It reframes something long treated as a private family matter as a public safety issue, on the grounds that the animal's situation is externally observable and is often the first thing to surface.

The law Wangwang never got may not ultimately be driven by animal protection at all, but by the more widely accepted argument that violence inside a home does not stay there. That is not the ideal path. It may be the first one that works.