First published February 2016. Rewritten July 2026 with India’s first national cruelty figures and ten years of FBI data.
In May 2026, India found out how many animal cruelty cases it registers in a year. The answer was 9,039.
Not 9,039 more than last year. Just 9,039, for the first time. The National Crime Records Bureau’s Crime in India 2024 is the first edition in the country’s history to give the Prevention of Cruelty to Animals Act, 1960 its own line in the national crime record. Before this, if you wanted to know how much animal cruelty India prosecutes, there was no number to look up. There was an argument.
Alongside the 9,039 cases: 10,312 people arrested, a 96.7% chargesheet rate and an 80.5% conviction rate, rising to 93.8% in metropolitan cities.
What 9,039 does not tell you
Look at the state split and the limitation becomes obvious. Maharashtra recorded 2,927 cases. Telangana 1,890. Kerala 1,510. Uttar Pradesh 1,121. Delhi recorded 35.
Nobody believes Delhi is eighty times kinder to animals than Maharashtra. What that gap measures is whether a police station will write an FIR when someone walks in about a poisoned street dog. This first count is a map of reporting behaviour, not of cruelty.
Which is fine. That is what a first count is for. Its value arrives next year, when there is something to compare it against, and the year after that, when a state’s number moving becomes a question somebody has to answer. We know this because another country already ran the experiment.
The FBI started counting ten years earlier
From 1 January 2016, the FBI began collecting animal cruelty as its own offence category in the National Incident-Based Reporting System, alongside homicide, arson and assault. Until then it had been filed under “all other offences”, a bucket that also holds spitting in public. The change came after roughly a decade of pushing by the Animal Welfare Institute and the National Sheriffs’ Association, with the animal welfare advocate Mary Lou Randour driving it.
They did one thing that India has not done: they split the category before collecting anything. Four sub-types from day one — simple or gross neglect, intentional abuse and torture, organised abuse such as fighting rings, and animal sexual abuse. Year one, with only 13 states reporting, produced 599 neglect incidents, 523 of intentional abuse, 12 organised, 9 sexual.
Ten years of that data is now in, and the analysis published by the Animal Welfare Institute in early 2026 shows what a split category buys you. Neglect cases are committed by men and women in roughly equal numbers. Intentional abuse is not: men are about four times more likely to be the person charged. Those are two different problems with two different interventions, and a single undifferentiated “cruelty” number would have hidden the difference entirely. The gap they still have not closed after a decade is hoarding, which has no sub-category and therefore no data.
India’s 9,039 is one undifferentiated number. Whether it stays that way is a decision somebody is making right now.
The serial killer story is the weakest version of this argument
This post used to open with Ted Bundy, Jeffrey Dahmer and David Berkowitz. Almost every article on this subject does. It is a compelling opening and it is the shakiest evidence in the field, so it is worth being straight about what it rests on.
The FBI profiler Robert Ressler’s line, that murderers “very often start out by killing and torturing animals as kids”, comes from work built on 36 sexual homicide offenders, with no control group and no statistical analysis. The broader idea it feeds, the MacDonald triad of animal cruelty, fire-setting and bed-wetting as a predictor of future killers, has been tested repeatedly since and does not hold up. Reviewers have found no compelling evidence that the triad predicts adult violent offending.
Two numbers that circulate constantly deserve the same treatment. The claim that a police study found 100% of sexual homicide offenders had a history of animal cruelty traces back to a newspaper mention of an unnamed study with no published sample size. And the widely quoted “70%” figure is real but routinely stated backwards. The 1997 Massachusetts SPCA and Northeastern University study followed 153 prosecuted animal abusers over twenty years against matched controls and found that 70% of the abusers went on to commit at least one other offence, with abusers about five times likelier than controls to commit a violent one. It does not say that 70% of violent criminals started with animals. Those are completely different claims, and the reversed version is the one that spreads.
Even the school-shooter statistic is softer than its reputation. The 2000 study behind it found animal abuse histories in five of eleven perpetrators, and the researchers who revisited it in 2014 did so specifically because the false-positive rate makes it useless as a screening tool. Millions of children pull the wings off insects. Almost none of them become dangerous adults.
None of which means the link is imaginary. It means the serial killer framing is the wrong place to look for it. The strong evidence is somewhere much more ordinary.
It shows up in the house, not on the manhunt
Writing in The Conversation in July 2026, Shane Bateman, a veterinary forensic scientist at the Ontario Veterinary College and a Humane Canada board member, sets out where the research has actually landed. Homes where intimate partner violence is happening show higher rates of animal maltreatment, and the link is with the more severe forms of it.
The mechanism is the part worth sitting with. Threatening a pet, injuring one, withholding veterinary treatment, abandoning the animal — these are not incidental cruelty happening near a violent person. They are instruments of coercive control, aimed at the bond rather than the animal. Hurting the dog is how you demonstrate what you are willing to do, and how you make leaving expensive.
It works. A study across sixteen Canadian domestic violence shelters found that concern for a pet’s safety shapes when a person decides to leave. Reviews of the wider literature put the share of survivors who delayed leaving, or returned, out of fear for an animal somewhere between 18% and 48%. The pet is not a sentimental complication in these cases. It is a lever.
Some jurisdictions have started writing that into law. Canada’s Protecting Victims Act received royal assent in June 2026, with most of it in force from 18 July 2026, and adds these behaviours to the Criminal Code as specific offences. Ontario’s Standing Committee on Justice Policy named animal maltreatment as a factor in intimate partner violence and child maltreatment in its December 2025 report.
Bateman’s sharpest point is not about law though. It is that a vet, a police constable, a child protection officer and a shelter worker can each be holding one piece of the same household’s story, and no system exists that puts the pieces on one table. Animal-related information almost never enters a formal violence risk assessment. The warning does not get missed because nobody saw it. It gets missed because the person who saw it had nowhere to put it.
What India still doesn’t count
Getting PCA cases into the NCRB record is real progress. It also makes three gaps visible that were easy to ignore while there was no data at all.
- Nothing connects the two files. The NCRB records a cruelty case as a crime head. It does not record whether the same household also has a domestic violence complaint. In India the violence link is not disputed so much as unmeasurable, because no dataset is built to show it.
- Nowhere for the animal to go. Countries that have worked on this for a decade have pet-inclusive shelter programmes, and even there roughly one in five domestic violence shelters accepts animals. India has no equivalent scheme, and no provision that lets a court keep a pet inside a protection order. A woman leaving with a dog is solving that problem alone, at 11pm, with no number to call.
- The penalty that everyone quotes is the wrong one. The PCA Act’s ₹50 fine for a first general offence is still unamended; a Parliamentary answer on 3 February 2026 confirmed the amendment bill remains under examination. But cruelty involving killing or maiming an animal now falls under Section 325 of the Bharatiya Nyaya Sanhita, which carries up to five years and, unlike the IPC sections it replaced, has no minimum animal-value threshold. That is the section that reaches street dogs, and most reporting still cites the ₹50.
We keep a full breakdown of what the statutes actually say in our guide to animal laws in India, including the ABC Rules and what an RWA can and cannot do.
If you see it happening
Advice on this topic tends to end at “speak up”, which is not advice. Concretely, in India:
- Photograph and timestamp before anything else. Injuries, the location, the animal’s condition. Almost every cruelty complaint that collapses does so on evidence, not on law.
- File under BNS 325, not only the PCA Act. Police stations turn people away over the ₹50 fine. A cognizable BNS section is harder to wave off.
- Bring an organisation with you. A complaint arriving with a rescue group attached moves at a different speed than one arriving alone. Start from your city’s animal helpline and shelter list.
- Take repeated, deliberate cruelty by a child seriously, without diagnosing anybody. Repeated, severe, remorseless harm to animals is worth an adult’s attention. One incident of a curious eight-year-old is not a prediction. The research does not support treating it as one.
- If there is a person being hurt in that house too, say so in the complaint. Nobody is going to join the two cases for you.
Cruelty toward animals in India is also shaped by how the state treats street animals generally, which is why the funding argument behind the Supreme Court’s stray dog verdict matters here, and why feeding strays keeps ending in police complaints against the feeder rather than the person throwing stones.
The FBI’s decision in 2016 was never really about animals. It was a bet that a category of harm nobody counted was carrying information about harm everybody cares about. Ten years of data says the bet was reasonable. India has now placed the same bet, quietly, in a table in a government PDF. What it is worth depends entirely on whether anyone asks the second-year question.