A seven-year-old in Tauranga watches his father shove his mother into the kitchen counter. Police are called. A family violence incident is logged. The court issues a protection order and refers the family for a child safety assessment — a psychologist's evaluation of how the violence has affected the kids, what support they need, whether it's safe for them to stay in the home.

That assessment used to happen almost automatically. This year, there's a decent chance it won't happen at all.

In 2024/2025, just 1,385 child safety assessments were completed nationwide — down from 1,880 the year before. That's a 26.3% drop in a single year. (Source: Ministry of Justice, family-violence — 3.Child safety programmes)

Five hundred fewer children got their assessment. Five hundred families where someone — a judge, a social worker, a psychologist — was supposed to check on the kids, figure out what damage had been done, and write a plan to keep them safe.

The Ministry of Justice doesn't explain why. It could be staffing. It could be funding. It could be that fewer cases are being referred in the first place — which raises its own uncomfortable questions about whether incidents are being downgraded or filtered out earlier in the system.

What we do know is this: child safety assessments exist because family violence doesn't just hurt the people being hit. Kids who witness it are at higher risk of developmental delays, behavioural problems, anxiety, depression. They're more likely to experience violence themselves as adults. The assessment is meant to catch that early — to intervene before the damage compounds.

When 500 fewer assessments happen, that's 500 families where nobody looked. Where a child who needed help might have been missed entirely. Where a parent who wanted to protect their kid didn't get the report they needed to keep the violent partner away.

This isn't a trend you can spin as good news. Fewer family violence incidents overall? That would be progress. Fewer assessments completed while incidents remain high? That's a system failing to follow through.

The assessments are supposed to be automatic in certain cases — when children are present during serious incidents, when protection orders are issued, when courts need evidence to make custody decisions. If those referrals are coming through but the assessments aren't happening, we've created a gap where the most vulnerable kids fall through.

And if the referrals themselves have dropped by a quarter, that's worse. It means someone along the chain — police, courts, social workers — decided those 500 cases didn't need assessing. That the risk wasn't high enough. That the kids would probably be fine.

Ask that seven-year-old in Tauranga if he's fine.