ICIAC 2011 Seminar stream 3B

Improving problem solving (G)

Abstracts and slides

Analyse2Advance – sharpening the Simple2Start methodology

Ian Bullen, Strategic Analytical Partnership Co-ordinator, Greater Manchester Against Crime

In 2010, Alan Edmunds’ Tilley Award winning Simple2Start methodology was rolled out across Greater Manchester as part of a multi-agency problem solving approach to crime reduction. Evaluation of the project identified reductions in Serious Acquisitive Crime levels over and above those seen in each of the 12 districts of Greater Manchester, indicating a successful implementation. However, the details beneath the collective figures showed that a number of the most problematic areas had not improved as well as the averages in each district and that, worse still, a handful of them had seen a notable increase in such offences.

The Analyse2Advance technique was developed as a response to this, aiming to fine-tune the methodology, more accurately targeting those areas requiring interventions. One of the accepted criticisms of Simple2Start is that, because street length is variable, there is an emphasis on longer streets and the focus of problem solving is diluted along the entire street length. In defence of this, the Simple2Start methodology concedes that offences do cluster together in specific locations along those streets (see Farrell & Pease in Wortley & Mazerrole, 2008). The Analyse2Advance technique aims to locate those clusters, which often form around streets rather than along them, often incorporating shorter streets that may well be overlooked using Simple2Start.

In principal the methodology is similar but the repeat street count of the initial work is replaced by using the Nearest Neighbour Hierarchical Clustering (NNh) technique available in Crimestat software. Using this software allows the analyst to set a minimum count level for each cluster at a sufficiently low level to provide a suitable number of clusters to fit into each of the tiered approaches set out in Simple2Start. This replaces the streets in the Simple2Start triangles with clusters and allows partners to prioritise clusters in the same way they had with streets. In addition, the ability to run Monte Carlo simulations against the data provides the analyst with information on the statistical significance of any clusters, allowing a more scientific separation of these tiers.

Testing the validity of the new methodology against historic data by comparing how both techniques fared showed that the Analyse2Advance technique identified a top tier of cluster areas that was four-times more accurate than the similar tier of repeat streets. By way of an example, analysis of Burglary Dwelling offences revealed that 40% occurred within just 1% of the total area of Greater Manchester, offering a real focus for intervention at a conurbation level and at a local level.

Multi-agency problem solving packs have been issued to all 12 Greater Manchester districts with key areas for Burglary Dwelling, Theft Of Motor Vehicle, Theft From Motor Vehicle, Personal Robbery and Anti-Social Behaviour identified for attention from partners in those areas. GMAC Champions have been tasked with managing these areas and the conurbation-wide Serious Acquisitive Crime group has been charged with monitoring progress in these areas. Evaluation of the initiative will follow in due course.

Presenter's slides: ICIAC11_3B_IBullen

Reducing burglaries in Enfield – Operation ‘Safe as Houses’

Iain Agar, Community Safety Partnership Analyst, London Borough of Enfield

In 2008/09 Enfield endured a 24% increase in burglary offending prompting concern from residents, local MP, local media and Government Office for London. By the end of the financial year Enfield had the 2nd highest rate in London and the 8th highest nationally. Analysis showed a very pronounced seasonal spike in the winter months. Furthermore, hotspot locations had remained the same for several years. Almost 1 in 5 offences occurred within just 4.6% of Enfield’s geographical area. Chronic hotspots were able to persist owing to an extensive network of alleyways. Analysis also uncovered low repeat victimisation, however, with the use of a crime analysis tool (near repeat calculator) we uncovered a very high incidence of near-repeats. Whilst there was an abundance of strong intelligence regarding victims and locations, there was limited intelligence on offenders. Very low sanctioned detections existed, of less than 15%. Arrests of notable offenders also had no impact on overall burglary levels. Emphasis on offenders had proved ineffective. It was decided a new approach to focus on victims and locations would be designed. The response focused on addressing the weaknesses identified on the victims and locations sides of the problem analysis triangle. Principles used included increasing the effort by offering locksmith services to properties in affected areas – target hardening. Controlled access to reduce opportunity in areas of rear-entry offending by implementing alley-gates. The deflection of offenders by giving away free security measures to residents in ‘hot streets’. Other supplementary activity to tackle ‘broken windows’ theory included graffiti and fly-tip removals and altering design to increase natural surveillance. LISA analysis methods were used to identify significant clusters and areas for intervention based on long term trends. Over 3,000 households were targeted for intervention in the most chronic hotspots whilst 88 alleygate schemes were implemented. There was a 46.7% reduction in targeted streets in the first year in comparison to a 7% decrease elsewhere in Enfield. Reductions in intervention areas have continued into June 2011 (reductions continue to be 8x greater than the borough average). A survey of residents (73% response rate) showed 93% satisfaction. One ward, Palmers Green, which had consistently been a top 3 volume ward between 2001-2009 had fallen to 9th in 2010-11. The change in the levels of offending in intervention areas (just 2.5% of Enfield housing stock) contributed to over 40% of Enfield’s burglary reduction in 2009/10 and over 70% in 2010/11.

Offending levels within the intervention area are at their lowest levels in over five years. At a cost of £231k, gained from a performance reward grant, the total social and economic costs saved as a result of reduced burglary in the first two years equates to £934k.

Presenter's slides: ICIAC11_3B_IAgar

Page last modified on 23 nov 11 15:40