Picture a social worker in Porirua sitting down with a seven-year-old whose father was convicted of domestic violence. The assessment takes weeks: interviews, home visits, safety plans. It's supposed to happen for every child exposed to family violence serious enough to land someone in court.

This year, that happened 495 fewer times than last year. (Source: Ministry of Justice, family-violence — 3.Child safety programmes)

The number of completed child safety assessments dropped from 1,880 in 2023/2024 to 1,385 in 2024/2025 — a 26.3% decline. That's not a minor blip. That's the biggest single-year drop in the dataset.

So what's happening? Two possibilities, and they point in opposite directions.

Option one: fewer children are being exposed to family violence serious enough to trigger these assessments. Fewer prosecutions, fewer convictions, fewer kids caught in the system. If that's the story, it's unambiguously good news — the kind of trend that should be shouted from the rooftops.

Option two: the same number of kids need these assessments, but the system isn't reaching them. Budget cuts, staff shortages, overloaded case workers. The assessments aren't happening because there's nobody to do them.

The Ministry of Justice data doesn't tell us which it is. It just shows the number dropping, sharply and suddenly, after years of relative stability.

Here's what makes this urgent: these assessments aren't bureaucratic box-ticking. They determine whether a child can safely stay in their home. They identify which kids need ongoing monitoring. They're the intervention that's supposed to break the cycle before another generation grows up watching violence.

When 495 fewer of them happen in a single year, that's either 495 kids who didn't need the help, or 495 kids who needed it and didn't get it.

The difference matters enormously. One version is a policy success. The other is children slipping through the cracks while politicians argue about crime statistics.

You'd think we'd know which one it is. You'd think someone would be asking. But the data sits there, showing a steep decline, and nobody's explaining it.

Nearly 1,400 children still went through these assessments this year. But the 500 who didn't — compared to last year — deserve more than silence.