Hutt, Oliver Kenneth; (2020) Understanding the deterrent effect of police patrol. Doctoral thesis (Ph.D), UCL (University College London).

Preview

Text OHutt Thesis Final.pdf - Accepted Version Download (7MB) | Preview

Abstract

The fact that crime clusters spatially has been known since at least the early 19th century. However, understanding of the extent and nature of this clustering at different areal units, and the fact that crime also clusters at different temporal scales is relatively new. Where previously the most at-risk areas (or `hot-spots') of crime were defined over areas the size of city districts and for periods of months if not years, the last decade has seen the focus shift to micro-places - areas of only a few hundred metres across - which are only `hot' for days or even hours. The notion that visible police presence in crime hot-spots can deter crime is not new and has been the basis of police patrols for two centuries. This deterrent effect has been well evidenced in many previous studies, both by academics and police practitioners. However, evaluations of these more recent micro-level hot-spot patrol strategies face significant analytic challenges and data quality concerns. They also often assume levels of police activity at the micro-area level (an `intention-to-treat' design) rather than measuring it directly. The aim of this thesis is to investigate the accuracy and precision of data that can be used to evaluate micro-level hot-spot patrol strategies and the implications this has for any analysis conducted using such data at these micro-level geographies. This thesis begins by outlining the relevant literature regarding place-based policing strategies and the current understanding of how crime clusters in both space and time. It continues by highlighting the data challenges associated with evaluating micro-level police interventions through the use of an illustrative analytic strategy before using a self-exciting point process model to evaluate the effects of police foot patrol in micro-level hot-spot under the assumption that the crime and patrol data being used are accurate. This is followed by two chapters which investigate the quality of the two datasets. Finally, the point-process evaluation is re-conducted using simulated data that takes account of the uncertainty of the datasets to demonstrate how data quality issues effect the result of such an evaluation and ultimately, the perceived efficacy of these highly-focussed policing strategies.

Download activity - last month

Export as CSVXMLJSON

Download activity - last 12 months

Export as JSONXMLCSV

Downloads by country - last 12 months

Export as XMLCSVJSON

Archive Staff Only

View Item