Skip to content

NZ Police / MoJ data via eolas

eolas serves 5 datasets from NZ Police + the Ministry of Justice — the canonical sources for road policing infringements and court charges in New Zealand.

If you're doing road-safety analysis, alcohol-policy research, court / criminal-justice work, or location-specific policing studies — this is the source.

For crashes (the consequence side), use NZTA / Waka Kotahi — they own CAS crash data. Police data covers enforcement actions (offences detected, charges filed).

What's in the catalogue

Road policing — NZ Police

Dataset Description
police_road_offences Monthly counts of road policing infringements by police district + area, 2009+.
police_camera_offences Monthly counts of camera-issued traffic offences by site location, 2009+.
police_breath_tests Quarterly passive + screening breath alcohol tests conducted by NZ Police district.

Court charges — Ministry of Justice

Dataset Description
moj_charges_by_offence Annual count of finalised court charges in NZ broken down by ANZSOC offence code.
moj_people_charged Annual count of distinct people with finalised court charges (defendant counts, not charge counts).

Refresh schedule

Quarterly. Police data updates monthly at source; MoJ data annually (calendar year basis, published mid-year following). Our refresh runs weekly to catch new releases promptly.

meta = client.info("police_road_offences")
meta["last_refreshed_at"]
meta["source_last_modified_at"]

License

All Police + MoJ data is published under CC-BY 4.0. Commercial use is fine; attribution required.

Recommended attribution: "Source: NZ Police" or "Source: Ministry of Justice", served via eolas (eolas.fyi). CC-BY 4.0.

A privacy note: data is aggregated. Small-cell suppression applies where individual case identification might be possible (typically counts ≤4 in specific demographic / location combinations).

Common patterns

Road-policing infringement trend by district

rp = client.police_moj("police_road_offences", start="2015-01-01")
# By district + month
by_dist = rp.groupby(["date", "police_district"])["count"].sum().reset_index()
pivot = by_dist.pivot(index="date", columns="police_district", values="count")
pivot.plot(title="Road policing infringements by district", figsize=(12, 6))

library(dplyr)
library(ggplot2)

rp <- eolas_get_police_moj("police_road_offences", start = "2015-01-01")
rp_by <- rp |> group_by(date, police_district) |> summarise(count = sum(count))
ggplot(rp_by, aes(date, count, colour = police_district)) + geom_line()

Camera offences by site

cam = client.police_moj("police_camera_offences", start="2020-01-01")
# Top 10 sites by total camera-issued offences
top = cam.groupby("site_location")["count"].sum().sort_values(ascending=False).head(10)
print(top)

Breath-test rate

bt = client.police_moj("police_breath_tests", start="2018-01-01")
# Positive-detection rate by quarter
bt["positive_rate"] = bt["positive_results"] / bt["total_tests"]
by_q = bt.groupby("date")["positive_rate"].mean()
by_q.plot(title="Positive breath-test detection rate")

MoJ charges by offence category

charges = client.police_moj("moj_charges_by_offence")
# Top offence categories most recent year
latest = charges[charges["year"] == charges["year"].max()]
print(latest.groupby("anzsoc_division")["charges"].sum().sort_values(ascending=False).head(10))

Source-specific notes

  • Infringements ≠ convictions ≠ crashes: a road-policing infringement is an offence detected by Police (driver issued a notice). It may or may not be paid / contested / withdrawn. For convictions, use MoJ data. For crashes, use NZTA CAS. Don't confuse the three.
  • Police districts vs council areas: NZ Police has 12 districts that don't align with regional or territorial council boundaries. Some districts (e.g. Counties Manukau) cover parts of multiple regions.
  • Camera-offences scope: includes red-light + speed cameras only. Some camera-issued offences (e.g. mobile cameras at specific events) may not be reflected.
  • Breath-test methodology change: pre-2020 vs post-2020 sampling methodology differs slightly. Compare cautiously across the boundary.
  • ANZSOC offence codes: Australian and New Zealand Standard Offence Classification — see stats.govt.nz/methods/anzsoc for the full taxonomy.
  • Defendant counts ≠ unique people: moj_people_charged is unique people per year. Someone charged in multiple years counts in each year separately.
  • Court charges are "finalised": i.e. the case has reached a final disposition (guilty / not guilty / withdrawn / etc.). Active / unresolved cases aren't in the data.

Where to find more