Predicting the future

Greater Manchester examines the next generation of strategic analysis

Strategic analysis in Greater Manchester has just been given a broader dimension as a result of a training course given to strategic analytical partnership coordinators employed by Greater Manchester Police as part of their Greater Manchester against Crime (GMAC) initiative.

The urban sociologist, planner and policy analyst, Dr Ned Levine, was invited to Manchester by the GMAC team to deliver training on the CrimeStat III spatial statistics program developed under his supervision and funded with a series of grants by the National Institute of Justice, Washington, DC.

CrimeStat has been developed over ten years and all versions (latest 3.1) are available free of charge. A manual/text book is provided on the CrimeStat website, which offers support and instructions to all users and it includes examples of how researchers can benefit from using CrimeStat.

The software has been developed with the aim of aiding: law enforcement practitioners and policy makers; criminal justice researchers; analysts; social scientists; epidemiologists; public health workers; and the forestry industry.

With the forum of law enforcement, CrimeStat facilitates the examination of crime hotspots, shifts in crime over time and patterns of serial offenders. The latter involves providing an estimation of the likely home location of the offenders as well as the temporal sequences of their committed crimes. CrimeStat links in with Geographic Information System (GIS) applications – in the case of Greater Manchester Police this is in the form of MapInfo tables. CrimeStat has the facility to output to ArcGIS, ArcView, MapInfo, Atlas*GIS Surfer for Windows, Spatial Analyst, Maptitude and many other desktop GIS packages. The input to CrimeStat is in the form of ‘dbf’, ‘shp’ and ASCII files. CrimeStat itself is not a GIS, it works with GIS systems to analyse crime patterns.

Preventing crime
“The purpose of CrimeStat and the training is to provide supplemental statistical tools to aid law enforcement agencies and criminal justice researchers in their crime mapping efforts,” said Dr Levine. “The National Institute of Justice is keen for this free program to be used by crime analysis practitioners in their efforts to understand the patterns and trends of crime and link them to a spatial context. It is envisaged that this intellectual system shall one day be able to predict crimes before they happen, in order for police to operate before the crimes take place.”

As explained and demonstrated in the training, Greater Manchester Police can use CrimeStat to aid crime prevention without spending large quantities of resources and funds on tests and projects. The system can be used to develop virtual models of strategies designed to reduce and prevent crime, which can be tested for their effectiveness before being put into practice. Geographic locations of hotspots, travel patterns of offenders and likely locations of future offences of serial offenders can be mapped and allow analysts and crime and disorder partners to develop strategies surrounding particular crimes or areas of disproportionate risk.

Whilst attending the training course, the analysts were taught Spatial Statistics routines (spatial distribution, distance analysis, and hotspot analysis), Spatial Modelling (interpolation, journey to crime analysis, crime travel demand modelling, Bayesian journey to crime analysis and space-time analysis), and Significance Testing (Monte Carlo simulation). The routines involved looking at the patterns and densities of crimes committed at conurbation level and locally with the emphasis on temporal and distance trends.

Analysis of the data projected on the maps reminded students that offenders are just the same as the rest of the population in that they will travel by the easiest routes according to their mode of travel; this should be taken into consideration when trying to determine the route taken by criminals when co