The Law That Couldn't Stop a Bite — and the Reform That Might

On 21 March 2026, Minister Simon Watts announced a complete review of the Dog Control Act 1996 — the first wholesale rewrite in the law's thirty-year history. The announcement came five weeks after Mihiata Te Rore was killed at Kaihu, Northland, in February 2026, in the fourth fatal dog attack in New Zealand in four years. A law that was already overdue for scrutiny had run out of excuses.

The Act was passed in 1996 when the internet was still a novelty and the breed-specific legislation debate was in its early stages globally. In the three decades since, it has banned five breeds, created two separate classifications of dangerous dog, and presided over a steady rise in dog bite hospitalisations. The reform announced in March 2026 is the most significant moment in the Act's history.

One law for a whole country

New Zealand's national architecture is one of its most valuable assets for reform. Australia has eight separate state and territory regimes for dog control — each with its own banned breeds list, its own classification system, its own enforcement approach. A change in Queensland tells you nothing about what happens in Victoria. New Zealand has one law. One Parliament, one Act, one reform process. What changes in Wellington changes everywhere. That is a significant advantage, provided the political will is present. In March 2026, it appears to be.

Five banned breeds — not four

The Schedule 4 breed ban under the Dog Control Act covers five breeds, not the four commonly cited in public discourse. American Pit Bull Terrier, Brazilian Fila, Dogo Argentino, and Japanese Tosa were banned in the original 1996 legislation. The Perro de Presa Canario was added by the Dog Control (Presa Canario) Order 2010 (SR 2010/369). At the time of the Presa Canario addition, the Department of Internal Affairs stated publicly that it was not aware of any Presa Canarios currently living in New Zealand. The breed was banned pre-emptively.

Menacing vs Dangerous: a tale of two classifications

The Act creates two post-ban classifications: 'menacing' and 'dangerous.' A menacing classification can be applied on the basis of breed or appearance alone — no incident is required. A dangerous classification requires a proven attack. The National Dog Database held 9,162 menacing dogs in the most recently available DIA figures, compared with 573 dangerous dogs — a ratio of approximately 16 to one. The overwhelming majority of classified dogs have never attacked anyone. They are classified because of how they look.

The bite data: what thirty years of ACC claims show

The Accident Compensation Corporation recorded 14,104 dog bite claims in 2022–23, at a total cost of NZD 15.8 million. The national rate was 2.75 claims per 1,000 people. Those figures represent an increase of approximately 44 percent in cost since 2017–18, when the equivalent figure was NZD 11 million.

A peer-reviewed study by Duncan-Sutherland and colleagues, published in the New Zealand Medical Journal in 2022, examined hospitalisation trends from 2014 to 2019 and found that the hospitalisation rate rose by 2.43 percent per year over that period. The pattern did not shift after the Presa Canario addition in 2010. Breed-specific additions had no measurable effect on the trajectory.

Four deaths in four years

Neville Thomson was killed in 2022. Elizabeth Whittaker was killed in 2023. Timothy Rolleston-Bryan, aged four, was killed in March 2025. Mihiata Te Rore was killed in February 2026. Four deaths in four years created political pressure that three decades of rising bite statistics had not. The pattern that distinguishes these cases from the long tail of non-fatal attacks is severity and age — the victims most at risk are children and older adults, attacked by dogs they knew, on properties they had every right to be on.

The criminal law has responded to at least one case. In R v Wira, Abel Jaye Wira became the first person in New Zealand to be convicted of manslaughter in connection with a fatal dog attack. He was sentenced to three years and six months' imprisonment — a precedent that signals the courts are willing to treat fatal dog attacks as serious criminal events, not mere accidents of ownership.

The science: what breed actually predicts

The scientific basis for breed-specific legislation has been eroding for over a decade. The landmark Morrill et al. study, published in Science in 2022 and based on 18,385 dogs and genetic data, found that breed explains only nine percent of behavioural variation in individual dogs. The SPCA New Zealand and the New Zealand Veterinary Association have both formally opposed automatic breed classification as a control mechanism, citing the scientific evidence. Classification by appearance — the current mechanism for the menacing designation — is particularly vulnerable to this critique: visual breed identification by non-experts is unreliable.

The SPCA open letter: nine bodies, one position

On 27 February 2026, SPCA New Zealand published an open letter on Dog Control Act reform, co-signed by the New Zealand Veterinary Association, Companion Animal Veterinarians, the New Zealand Animal Law Association, and five other organisations — nine bodies in total. The letter called for a complete rewrite of the Act, the removal of breed-specific restrictions, the introduction of mandatory desexing and mandatory injury notification, and minimum fencing and housing standards.

What the 2026 review proposes

Of the proposals under consideration, mandatory desexing has the strongest evidence base. The TeamMate study of New Zealand working dogs found only six percent were neutered — versus approximately 85 percent of pet dogs. Intact male dogs are over-represented in serious attack statistics in every jurisdiction where data has been collected. Desexing is not a complete solution, but it is among the most evidence-supported single interventions available.

Mandatory injury notification closes a data gap that has persisted for the full thirty-year life of the Act. Currently, there is no requirement for hospitals or emergency departments to notify local councils when a patient is admitted for a dog bite injury. Councils have no systematic picture of the incidents on their patch. Without that data, it is impossible to identify patterns, target enforcement, or evaluate whether interventions are working.

The road from here

Thirty years of evidence points in a consistent direction. Breed alone is a poor predictor of individual dog behaviour. The highest-risk environments are private properties, not public parks. The most vulnerable victims are children and older adults. The most effective single interventions are desexing and owner accountability. New Zealand has the architecture — one law, one Parliament, one reform moment — to rewrite once and apply everywhere. The review announced in March 2026 is the most significant opportunity this country has had to get dog control right in a generation.