Methodology

Data source

Suburb Sense uses NZ Police Recorded Crime Victim Statistics (RCVS) from policedata.nz. Briefs are built from the Victimisations report with location and time detail (Area Unit, meshblock, day, and hour) — not demographics-only extracts. Population and area hierarchies come from Stats NZ geographic tables and 2018 census meshblock population.

How numbers are produced

A batch data pipeline ingests the Police extract, cleans and validates fields, joins geography and population, then aggregates to Area Unit (suburb-scale) statistics. Results are published as static JSON for this site. Rates use population as a stable baseline; confidence rules reduce emphasis on rates when counts or population are small.

Comparison windows

  • Rolling 12 months — latest data month back 12 months vs the prior 12 months (primary comparison).
  • Latest month — most recent month vs the month before (shown only when confidence is sufficient).
  • Longer-term trend — rolling 12-month rate per 1,000 over time. The first chart point appears only after a full year of history is available in the source extract.

Rates per 1,000

Rates divide reported victimisation counts by population. We use 2018 census population rolled to 2017 Area Units as a stable baseline — they show relative patterns, not exact current risk.

Geography

Statistics are shown at Area Unit (~suburb) level. Boundaries may not match colloquial suburb names or individual streets.

Limitations

  • Not all crime is reported to Police.
  • Dwelling-based victimisations may be excluded from some Police public views.
  • Time of day is often unknown for property crime.
  • Calendar week trends are not available — Police data uses month, day-of-week, and hour.

Glossary