Heterogeneous Crime Disruption in the COVID-19 Era: Structural Break Evidence From Aotearoa New Zealand Monthly Victimisation Data

This study examines whether the COVID-19 period coincides with structural breaks in Aotearoa New Zealand recorded crime and whether break timing differs across offence categories. Monthly report-date victimisation counts from the Aotearoa New Zealand Police Recorded Crime Victim Statistics are analysed from July 2014 to January 2025 across nine offence groupings. Log-transformed series are modelled with seasonal controls and long-run trend terms. Structural change is assessed with the Bai-Perron endogenous break test, which searches all candidate months for the dates at which the level or trend of a series changes, using wild-bootstrap inference and information-criterion selection of multiple breaks; corroborating stability diagnostics are reported as sensitivity analyses. Magnitudes are characterised through a centred pre–post regime comparison with heteroskedasticity and autocorrelation-consistent inference, and a decomposition separates persistent change from recovery after the restriction period. The results indicate widespread but non-universal instability and reveal three timing pathways: near-onset breaks for opportunity-sensitive property offences, pre-pandemic breaks for selected violence-related categories consistent with shifting baselines, and delayed breaks in late 2021 to mid-2022 that are consistent with, though not uniquely attributable to, adaptation and digitalisation. Overall, the findings support a multi-phase interpretation of pandemic-era crime change and underscore the importance of aligning inference about disruption mechanisms with offence-specific opportunity structures and measurement conditions in recorded crime.