Data & wrangling

• Source: NSW BOCSAR, LGA monthly incidents (1995–2025)
  
• Wrangle: clean_names() → pivot_longer() → parse month with lubridate::my()
  
• Focus: Theft → Steal from motor vehicle; LGAs: Sydney, Parramatta, Canterbury-Bankstown, Randwick
  
• Methods: 3-month moving average; latest-year totals; LOESS trend
  
• Limits: Counts (no population rates); possible reporting changes over time

Visual 1 — 3-month moving average

Visual 2 — Top LGAs (latest full year)

Visual 3 — Local trajectories (4 LGAs)

Limits & next steps

• Counts, not rates — no population adjustment; 3-month MA smooths noise but can blur short spikes
• Comparability — recording/policy/practice changes across years may affect levels; possible missing/suppressed values
• Next — add rates per 100k + SEIFA/vehicle density for context; publish code/data link for reproducibility

References (APA 7th)

• NSW Bureau of Crime Statistics and Research. (2025). Recorded crime statistics: Criminal incidents by local government area, monthly, 1995–present [Dataset].
  Retrieved from https://bocsar.nsw.gov.au/statistics-dashboards/crime-and-policing/lga-excel-crime-tables.html

• NSW Bureau of Crime Statistics and Research. (2025). Criminal offences open datasets [Dataset].
  Retrieved from https://bocsar.nsw.gov.au/statistics-dashboards/open-datasets/criminal-offences-data.html

• NSW Government Data Portal. (2025). Crime data by offence – NSW [Dataset].
  Retrieved from https://data.nsw.gov.au/data/dataset/crime-data-by-offence