Author Joanna JJ Wang, Don Weatherburn and Wai-Yin Wan
Published November 2019
Report Type Crime and Justice Bulletin No. 225
Subject Assault; Statistical methods and modelling; Vandalism and property damage; Evaluation reports; Policing
Keywords Police activity, violent crime, property crime, panel cointegration, pooled mean group estimator

Download this publication



To examine the long-term effect of two routine police activities on property and violent crime in NSW.


Police move-on directions and person searches as well as property and violent crimes were extracted from the NSW Police Force’s Computerised Operational Policing System. We investigate the relationship between police activity and crime using panel of 17 Local Area Commands (LACs) over the period 2001 to 2013. To estimate the long-run relationship, panel models of Pooled Mean Group and Mean Group were applied to allow for differing effects between LACs.


We estimated a significant and strongly negative long-run relationship between both indices of police activity (move-on directions and person searches) and each of break and enter, motor vehicle theft and robbery. The person search activity is negatively related to assault, but the effect is weak; with a 10 per cent increase in person searches only producing a 0.29 per cent fall in assaults. No significant long-run relationship was found between assault and move-on directions.


Sustained increases in police activity, whether in the form of move-on directions or person searches, do appear to help suppress break and enter, motor vehicle theft and robbery but do not appear to help in reducing assault.

Download this publication