What happens to a charge that never reaches a verdict?

Sometimes it just stops. The prosecution pulls it, the file closes, and the case leaves the system without a conviction, an acquittal or a discharge. Stats NZ counts those outcomes as "withdrawn", and in 2025 there were 42,311 of them across all offence types and all court regions (Source: Stats NZ, charges-by-offence-type). That is the highest annual figure since 2010, when 48,662 charges were withdrawn.

Rewind a decade and the picture is very different. Withdrawals bottomed out at 28,120 in 2015, the lowest point in the entire series going back to 1986. They stayed near that level through 2016, drifted up into the low 30,000s, then dipped again to 30,120 in 2021. Since then they have risen every single year: 33,057, then 37,369, then 41,679, then 42,311. That is a 40% increase in four years.

Here is the detail I think matters more than the headline total. The four-year climb is losing steam. The jump from 2023 to 2024 was 4,310 charges. The jump from 2024 to 2025 was 632, an increase of about 1.5%. If withdrawals were on a runaway trajectory, you would not expect the newest year to be the flattest of the run. The most likely reading is that the post-2021 surge is levelling off near the low 40,000s rather than continuing upward.

The long view also punctures any sense that 42,311 is unprecedented. Withdrawals sat above 48,000 in every year from 1996 through to 2010, peaking at 58,396 in 2009. Then they halved. The fall from that 2009 peak to the 2015 floor wiped out more than 30,000 withdrawn charges a year. What we are watching now is a partial recovery toward mid-1980s levels, not a new record. The 1987 figure of 50,373 is still well clear of anything in the last 15 years.

The honest caveat: this dataset counts charges, not rates. If police lay more charges overall, withdrawals will rise even if the share of cases dropped stays identical. Without the denominator sitting alongside it, you cannot say from this table alone that prosecutors have become more likely to abandon a case. A four-year rise in raw withdrawals is consistent with more charging activity, with slower courts, with changes in charging thresholds, or with all three at once. Anyone telling you it proves one specific thing is going further than the numbers allow.

What the table does establish is scale, and scale is the thing that gets skipped. Roughly 42,000 charges a year end without any judicial finding at all. Every one of those involved someone accused, usually someone who appeared in court, and in many cases someone who waited months for an outcome that never came.

My view: the crackdown conversation in this country is almost entirely about what happens after conviction. The quieter number is how many charges never get that far, and it has now grown for four consecutive years. That deserves the same scrutiny as sentencing lengths, because it is measuring whether the front end of the system is laying charges that stick.