The search for five missing New Zealanders in Nepal has held the country's attention this week, with RNZ reporting on the flash flood near the Nepal-China border and the guide who lost multiple family members. It is the kind of risk that stops you cold: sudden, far away, unimaginable.

Here is the number that puts the risk closer to home. In 2024, the most recent year available, New Zealand recorded 14,283 serious non-fatal injuries across all causes. That is roughly 39 people a day left with an injury severe enough to meet the official serious threshold. (Source: Stats NZ / ACC, fatal-serious-injuries)

In 2000, the figure was 7,650. The count has almost doubled in a generation.

The shape of that rise is the interesting part. Through the 2000s, the number barely moved. It sat between 7,410 and 7,815 every year from 2002 to 2008, drifting up and down without direction. Then something changed. From 2009 onward, the count rose every single year. Not once in sixteen consecutive years has New Zealand posted a lower serious injury total than the year before.

The steepest stretch was the mid-2010s. Between 2013 and 2019, serious non-fatal injuries went from 9,117 to 12,945, a 42 percent jump in six years. The single year from 2018 to 2019 added 960 injuries on its own.

Now the good news, because it is real. Since 2019 the climb has almost stopped. The last five years added 1,338 serious injuries in total, a rise of about 10 percent. Compare that with the four years to 2019, which added 3,084. The most recent year, 2023 to 2024, added 210, an increase of 1.5 percent. The line is still pointing the wrong way, but it has gone from a steep climb to something close to a plateau.

The honest alternative reading: this is a raw count, not a rate. New Zealand's population grew substantially over these 24 years and got older, and both of those push injury counts up regardless of how safe the country actually is. A flat count against a growing population would mean the underlying rate is falling. The 2020 and 2021 figures also cover years when lockdowns cut driving, sport and travel, which may make the post-2019 flattening look better than the underlying trend deserves. And "serious" is a coding threshold, so changes in hospital recording practice can move the number without anything changing on the ground.

Even allowing for all of that, one fact survives. Sixteen years without a single annual decline is not what a country trying to get safer looks like. Falls, crashes, workplace incidents and assaults all feed this number, and every unit in it is a person who went to hospital and probably did not come out the same.

My view: the plateau since 2019 is the first genuinely encouraging signal in this series since the 2000s, and it deserves more attention than it gets. But a plateau at 39 people a day is a strange thing to celebrate. The target is a year that goes down. We have not had one since 2008.