Part III-C Learning from Practice: GIS as a Tool in Planning Sustainable Development Society and Environment © 2006 by Taylor & Francis Group, LLC 385 22 A Geographical Approach to Community Safety: A U.K. Perspective Jonathan Corcoran and Bernadette Bowen Thomson CONTENTS 22.1 Introduction 385 22.2 The Importance of Geography 387 22.2.1 What is Community Safety? 387 22.2.2 Current Approaches to Community Safety 388 22.3 The HASCADE Approach to Community Safety 388 22.3.1 Data Requirements and Issues 389 22.3.1.1 Technical Issues 389 22.3.1.2 Security Issues 389 22.3.1.3 Data issues 390 22.4 The HASCADE Model 390 22.4.1 Spatial Methods 391 22.4.2 Statistical Analysis 392 22.4.3 Results from HASCADE 393 22.5 Discussion 395 22.6 Future Developments 399 22.6.1 An Integrated Deployable Solution 399 22.6.2 Increased Data Sets 399 References 400 22.1 INTRODUCTION Crime and disorder are inevitable realities of society, affecting all of the populace either directly or indirectly. Their formal control has traditionally been the respon- sibility of the police. Increasingly, recent years have seen the control of crime and disorder, in England and Wales, charged to a range of both nationally and locally governed agencies. The requirement to minimize community problems through tackling crime and disorder issues was formalized in the Crime and Disorder Act [1]. The Act formally introduces the creation of multiagency Crime and Disorder Reduction Partnerships (CDRPs) within each local authority area. A legal obligation © 2006 by Taylor & Francis Group, LLC 386 GIS for Sustainable Development was placed on these CDRPs, particularly the local authority and police, to work in tandem to develop, publish, and implement three-year strategies to tackle crime and disorder. The production of an informed crime and disorder strategy relies heavily upon an in-depth local community safety audit, which provides a snapshot of crime and disorder-related issues, a further stipulation of the Act. Each audit, consisting of multiagency data and community consultation, attempts to encapsulate the com- munity dynamics within a given area. In addition, the Act stipulates the necessity to work with other key agencies, including the health authority (Sections 5–7, Crime and Disorder Act 1998 [1]), while Guidance recommends the expansion of the partnership to business and voluntary sectors. Section 17 of the 1998 Crime and Disorder Act extends the scope of responsibility for controlling crime and disorder. It places a statutory obligation on local authorities and the police to consider crime and disorder implications in all its functions [1]. Part of this legislation (Section 115) enables partners to share previously internalized data for crime and disorder reduction purposes. If these agencies are to embrace the principles within the Act, then the production of a holistic strategy is essential. Such a strategy would enable a variety of agencies to use their expertise for crime and disorder reduction and prevention purposes and for the benefit of the community, thus realizing increased community safety. Key to achieving their missions is the ability to assimilate an under- standing of criminal dynamics, which are inherently complex. Geographical tools have the potential to provide invaluable insight into these dynamics. It has been shown that crime and disorder recorded by the police constitute only a partial descriptor of community issues [2]. Therefore, to understand the dynamics and requirements of a region, there is the need to consult additional data, sourced from a range of organizations at the local level [3]. On this basis, local partnerships have been promoted to guide and facilitate the data collation, aggregation, and analysis process. Hough and Tilley [4] outline six guiding principles that support the requirement for local partnerships: • The police alone cannot control crime and disorder • No single agency can control crime and disorder • Agencies with a contribution to reducing crime and disorder need to work in partnership • Evidence-based problem solving approaches promise the most effective approach to reducing crime and disorder • Problems of crime and disorder are complex, and there are therefore no panaceas • Crime and disorder problems need to be understood in their local contexts and strategies need thus to be locally tailored Hough and Tilley [4, p.1]. With permission. In the remainder of this chapter, the importance of geography for crime and disorder analysis and nature of community safety is discussed. This is followed by a discus- sion of the design, development, and implementation of Holistic approach to strategic crime and disorder evaluation (HASCADE), a geographical approach to strategic crime and disorder analysis. © 2006 by Taylor & Francis Group, LLC A Geographical Approach to Community Safety: A U.K. Perspective 387 22.2 THE IMPORTANCE OF GEOGRAPHY The mapping of crime has a long history as a tool for understanding crime’s spatial distributions. It can be traced back as far as the nineteenth century in France [5,6], where mapping was first utilized to visualize and analyze crime information. Crime data and the computational tools that are used for their collection and analysis have, over recent years, grown in importance. Academics and practitioners have seen value in their potential to analyze crime and disorder issues. A central theme in the geo-analysis of crime and disorder data is the quest to better understand their dynamics, which in turn can be applied to formulate targeted responses. The U.K. Home Office advocates a geographically orientated approach to crime analysis. This is reflected by the marked growth of computerized mapping by U.K. police forces [7], the trend set to continue. However, a report of the auditing process [8] revealed that less than half (42%) made use of a GIS. 22.2.1 W HAT IS C OMMUNITY S AFETY ? Community safety is a recent concept, the definition of which has amassed much debate. Since the Morgan Report [9], the term community safety has witnessed increasing popularity in Britain. The Morgan Report (para. 3.7), considers commu- nity safety “as being concerned with people, communities and organizations includ- ing families, victims and at risk groups, as well as attempting to reduce particular types of crime and the fear of crime. Community safety should be seen as the legitimate concern of all in the local community” (cited [10, p.6]). Community safety is recognized as comprising situational and social characteristics. The situational characteristics of community safety include crime prevention. Crime prevention, in its simplest form, indicates a situation whereby crimes would have occurred if they had not been prevented [11]. In general, crime prevention techniques can be applied to a variety of approaches that aim to reduce the likelihood of an individual or group encountering crime events. Social characteristics refer to the socioeconomic and cultural aspects of people’s lives; thus individuals and groups should be able “to pursue, and obtain fullest benefits from, their social and economic lives without fear community safety because it recognizes that community safety is not the sole responsibility of the police. The change in policy focus toward partnership working at the community level in the United Kingdom implies that an advantage will be achieved if broader multiagency and multifaceted approaches are applied. Walklate identifies that “a genuine desire for policy to work for change needs above all to be cognizant of the importance of the local context in which that policy is set. This desire needs to work with rather than against the historical and socioeconomic circumstances which structure that local context” [13, p.62]. As such, diligence in ensuring that inclusive approaches are implemented and that these approaches are appropriate to the community should have primacy. HASCADE attempts to inform short, medium, and long-term strategy through the examination of data at the community level. This information endeavors to provide insight into the crime, disorder, and potential vulnerabilities that include © 2006 by Taylor & Francis Group, LLC or hindrance from crime and disorder” [12]. Partnership working is fundamental to 388 GIS for Sustainable Development socioeconomic factors present within such geographical areas. It seems, then, that the term community safety renders itself more easily toward applying holistic approaches, thus potentially increasing engagement from partners, community, and agency. In relation to this chapter, the term community safety will be used, recog- nizing that such a term includes crime prevention and that its definition can extend beyond the realms of crime and disorder. 22.2.2 C URRENT A PPROACHES TO C OMMUNITY S AFETY Current approaches to achieving community safety in England and Wales often reflect traditional crime prevention concepts, commonly involving applied situational crime prevention (SCP) techniques [14]. The application of such techniques has positively impacted upon crime and disorder reduction in communities, often achiev- ing a rapid effect. Felson and Clarke [15] note numerous examples where the application of targeted “opportunity-reducing measures” has produced effective out- comes. Such measures include reductions in check frauds occurring in Sweden, through the introduction of new identification measures, and the establishment of CCTV cameras in Surrey University car parks that resulted in reductions in crime. The HASCADE model is receptive to the important contribution of crime pre- vention, but it also endeavors to inform strategy development centered upon con- sideration of wider, holistic, community safety issues, particularly those based around social exclusion. Crime and disorder strategies informed solely by analyzing the spatial distribution of crime and disorder events, provide only a partial view of community issues. Typically this can involve the analysis of police crime and disorder data, to identify hotspots (areas exhibiting disproportionately high levels of crime and disorder). The results from this exercise are then used to design strategy to combat crime and disorder within the identified locales. However, such use of crime and disorder data is likely to increase the risk of only responding to community safety through the application of primarily situational methods (for example, the use of locks to deter burglaries within identified burglary hotspots). In addition, these are likely to result in imbalanced strategy that may not identify key facets of a community’s needs. Such imbalanced approaches increase the risk of exclusion, while potentially reducing trust within communities (for a more detailed discussion HASCADE introduces a joined-up strategic framework. Its holistic nature encourages multifaceted methods for improving community safety, as opposed to applying a singular methodology. The use of geographically referenced multiagency data is key to informing this holistic approach. 22.3 THE HASCADE APPROACH TO COMMUNITY SAFETY The relevance of incorporating multiagency data within the crime and disorder audits reinforces the underlying principles of partnership working that is promoted throughout the Crime and Disorder Act. In addition, all guidance relating to the crime and disorder audits has advocated the use of multiagency data sets [4], but little information exists yet on how these multiple data sets should be incorporated and analyzed. © 2006 by Taylor & Francis Group, LLC of trust and exclusion see [16,17]). A Geographical Approach to Community Safety: A U.K. Perspective 389 22.3.1 D ATA R EQUIREMENTS AND I SSUES The importance of a geographically orientated approach has already been stated. However, modeling the geography of crime in a way applicable to CDRP objectives requires an alternative approach to that demanded by police operations. Modeling techniques to direct, monitor, and evaluate community initiatives demands the adop- tion of a holistic approach, in which a range of local information is analyzed in an appropriate manner. The ability to visualize the precise locations of events has been welcomed and promoted by the government [18]. As such, microlevel analysis has become of particular interest for those implementing SCP programs. Microlevel analysis can prove successful in such programs, because the objective is to uncover the specifics of a locale. Furthermore, it offers an explanation of a locale’s propensity toward observed events (for example, a series of houses within a neighborhood particularly subject to burglary). SCP techniques should not constitute the entire audit analysis, because an imbalanced strategy, with a tendency to short-term gains, would result. Such a strategy would discount the social, cultural, and community characteristics of an area. A primarily SCP focus impedes the development of a strategy that strives to address community safety in the short, medium, and long term. Moreover, in the context of the audit, the use of such techniques places large demands upon each partner to provide full address information from which the data can be geocoded to the fine scale typically demanded by SCP programs. In addition, the role of the audit is to provide an overview of a whole local authority area; thus fine resolution analysis is arguably not the primary objective. Therefore, the audit should put in place a series of analyses that are capable of identifying the broad issues. At this stage a micro level analysis could take place to isolate the specific issues (for example, vulnerable houses and common modus operandi) to ensure a correct application of preventative measures (such as a lock-fitting scheme and security advice), while simultaneously directing attention on the broader social issues that impact upon the community. During the development of HASCADE, several constraining political, technical, and administrative issues were identified. These issues are discussed below in turn. 22.3.1.1 Technical Issues A fundamental issue at the commencement of auditing was the identification of key technical personnel within each partner agency, possessing the necessary technical skills and knowledge pertaining to their data systems. This formed a vital stage, following which questions regarding data (for example, descriptions of coding pro- tocols) could be posed and replied to efficiently. 22.3.1.2 Security Issues Despite the caveat (Section 115) in the 1998 Crime and Disorder Act facilitating the sharing of previously internalized data, many concerns for partners still remained. This was partially resolved through the implementation of data security measures in addition to the specification of what could be disseminated. Clearly stated dissemination © 2006 by Taylor & Francis Group, LLC 390 GIS for Sustainable Development protocols were enforced, whereby no information was to be published (cartographic or text based) beyond the confines of the CDRP without formal approval by all partners. On a technical level, it was agreed that a single machine would be utilized for all data analysis and presentation, the access to which was strictly controlled. 22.3.1.3 Data issues Many partners had concerns regarding the sharing of data. The concern was in perceived contravention of the 1998 Data Protection Act, despite the clear caveat stipulated within Section 115 of the Crime and Disorder Act facilitating their use. To appease such concerns, a data wish list proforma was designed to adhere to the general, while providing the necessary information from which targeted mapping and statistical output could be generated. Using the identified technical contacts, all partners completed this, and a final standard was agreed, with each agency providing as a minimum the contents summarized in Table 22.1. In addition, the confirmation of any problems, inconsistencies, and known errors were established prior to the analysis, which in the main involved changes in counting rules and coding protocols. This established known, but not necessarily published, information concerning data reliability. The result of this process was to either omit or amend their use within the audit. This could then be attached to the audit, not necessarily as a formal appendix, but as a reference from which decisions to include or omit certain facets of data could be supported. One of the most problematic issues was achieving a common temporal coverage across all contributing agencies. Typical obstacles to accomplishing this were mod- ifications to software systems that rendered data prior to particular dates difficult to access. In addition, the requested data spanned alterations in collection protocols and, thus, introduced potentially immeasurable inconsistencies. 22.4 THE HASCADE MODEL provide an insight into the dynamics of crime, disorder, and vulnerability across the partnership region. The result of the data collection process (following the resolution of the aforemen- tioned issues), achieved the collection of eight partners’ data spanning a twelve-month TABLE 22.1 Minimum Data Provision Proforma Attribute Content Temporal Reference Date (dd/mm/yyyy) Spatial Reference Full postcode Incident or Event description Numeric code or standardized text File formats MS Excel, MS Access, Delimited text © 2006 by Taylor & Francis Group, LLC The HASCADE model (Figure 22.1) uses both spatial and statistical techniques to A Geographical Approach to Community Safety: A U.K. Perspective 391 period. Each data set was then attributed a meaning, classified as either a crime and disorder data set or a vulnerability data set (Table 22.2). The following sections detail the spatial and statistical techniques that were employed to generate the final output. 22.4.1 SPATIAL METHODS Using the spatial reference provided by all partners, data was geocoded utilizing OS Code-Point ® [19] to assign x,y coordinates at a postcode unit level. Once geo- referenced, all data were inspected at point level and overlain with street and boundary information to provide context. In many cases the volume of mapped incidents created visualization problems at the point level, where multiple incident localities appeared as a single occurrence (each point was simply positioned one above another and, thus, appeared incorrectly as a single incident). Therefore aggregate and density FIGURE 22.1 The HASCADE model. TABLE 22.2 Meanings Attributed to Data Crime and Vulnerability Indicators Associated Datasets Crime and Disorder Police Incident Figures, Benefit Fraud, Arson Vulnerability indicators Economic poverty Council tax benefit claimants Peer and family criminality Police reprimand, final warning, sentence — youth offending team (YOT); supervised and unsupervised data — probation Lack of educational attachment and future risk of low educational attainment School exclusions Risk of social exclusion All of the above, looked after children and benefit fraud DATA COLLATION/ INTEGRATION DATA ANALYSIS/ (Driven by criminological theory) OUTPUT GIS Spatial interrogation Statistical Underlying correlations Identifications of PGAs Determination of areas in FACTORS A,B,C,D © 2006 by Taylor & Francis Group, LLC 392 GIS for Sustainable Development mapping were used to provide a better indication of event intensity across the CDRP region. In order to provide a greater context to the underlying population geography to which the various events were related, it was necessary to generate a series of aggregate maps. Using a boundary set based upon the 1991 Census enumeration districts (containing 1999 population estimates) each partner’s data were aggregated to the new framework. A GIS script was created to automate the calculation of number of incidents contained within each region and derivation of rate based upon population. The script consisted of a “point in polygon” test for the partner’s data to calculate the total number of incidents occurring in each region. The total count was then used against the population for that region to derive the incident rate per 1000 population. A key part of the spatial analysis was to derive a boundary network to provide the closest representation of each partner’s data. This boundary network then formed the foundation from which statistical analysis could be conducted. Creating and validating the boundary network first involved examining the event distributions (using hotspot mapping — as this best describes event distribution) from each agency’s data. Where there was an identified lack of coterminosity, the boundary network could be modified to provide a closer fit to the agency data. Typically, a modification included aggregating two or more regions together. In certain circum- stances, however, the imposed boundaries could either under- or overfit areas of high event volume. Because a perfect match could not be achieved, it was deemed acceptable One limitation of this process was the input boundary network constraining the minimum size of areas for which population data was available. Thus, if agency data underfitted a region, this could not be redefined to a subdivision, because incident rates could not be calculated. A second limitation was the decision criteria used to assess whether areas were representative or required modification. For this stage visual inspection rather than any quantitative techniques were employed because it was recognized that high-level precision could not be achieved due to the nature of the agency’s data. A visual comparison was therefore considered sufficient. At this stage it was possible to commence a primary identification of priority geographical areas (PGAs) on the strength of point, hotspot, and aggregate mapping across all data sets. Aggregate outputs offered an indication of vulnerability at a broad neighborhood scale, while point and hotspot maps identify more specific subneighborhood localities internal to these regions. 22.4.2 S TATISTICAL A NALYSIS Spatial analysis using GIS techniques offered a tool by which visualization and aggregation was conducted. In many audits, outputs from this stage are taken no further. Therefore one was able to overlay and visualize, but unable to quantify interactions between various layers of information. Statistical analysis targeted the correlation between the various data sets to reinforce relationships identified through the spatial analyses. Establishing significant statistical linkages, together with identification of PGAs, provided the foundation from which © 2006 by Taylor & Francis Group, LLC where the boundaries were generally representative of all data, (Figure 22.2). A Geographical Approach to Community Safety: A U.K. Perspective 393 partnership strategy was developed through building a fundamental comprehension of community safety processes. Pearson correlation coefficients [20] were used as the basis from which signif- icant relationships were established. Using the rates for each data set, for each small area, significant correlations were flagged and used as the basis to establish key dependencies between the various data areas (the term small area refers to a single areal unit of the validated boundary network). Thus, for a neighborhood area iden- tified as a vulnerable locale for events such as school exclusions and youth offending, spatial inspection can be further queried through statistical linkages, in addition to suggestions of further potential associations that may exist. The result of this hypo- thetical scenario would be to identify a target set of agencies required to jointly direct interventions in the specified area. In addition, such output and evidence produced through the implementation of this framework reinforces the necessity of the partnership, supporting the current drive toward joined-up government working practices. 22.4.3 R ESULTS FROM HASCADE The final part of the analysis was to produce a final output in which areas from the validated boundary were classified to reflect each of the partners’ data. To achieve this categorization, a count system was designed. Using a natural break method, each partner’s data was first mapped using five classes. For each data set, those areas FIGURE 22.2 Method of validating the boundary network. Identification of areas with sim- ilar event intensities was achieved using the GIS to overlay each agency’s data (for example, see dotted line between layers). The degree to which hotspots matched the boundary network could be assessed and modifications made if the majority of layers lacked coterminosity. (From Corcoran, J. and Bowen Thomson, B., Br. J. Commun. Justice, 2 (1), 45, 2003. With permission.) © 2006 by Taylor & Francis Group, LLC that were classified in the top two categories were tabulated (Figure 22.3). [...]... 23.2 as TO4-A, archeological areas; TO4-B, archeological potentiality areas; TO4-C, delimited archeological areas; TO4-D, isolated emergencies; and TO4-E, archeo-termal areas © 2006 by Taylor & Francis Group, LLC GIS for Sustainable Development Rules GIS Application to Support Water Infrastructures Facilities Localization 413 FIGURE 23.4 Landscape Plan suitability for water infrastructures for Stromboli... relevance also as informative support to start up a participatory process This appears particularly advisable in contexts, © 2006 by Taylor & Francis Group, LLC 414 GIS for Sustainable Development FIGURE 23.5 Land use suitability for water infrastructures for Stromboli Island FIGURE 23.6 Macro-location suitability of water infrastructure for Stromboli Island © 2006 by Taylor & Francis Group, LLC GIS Application... Guidance for CDRPs notes the importance of referring to multiagency information to inform strategy The partnership Auditing Toolkit [21] advocates the use of GIS and its incorporation into crime and disorder audits, but its emphasis is upon the use of such systems for microlevel analysis of crime and disorder incidents, thus risking © 2006 by Taylor & Francis Group, LLC 396 GIS for Sustainable Development. .. regarding the initial provision of information is increased for agencies that may be hesitant about sharing personalized information Future developments for HASCADE would include expanding the crime, disorder, and vulnerability data layers currently incorporated in the model The minimum data requirements increase the potential for agencies, particularly smaller, or not -for- profit agencies, to be included... illustrating) 22.6.2 INCREASED DATA SETS In spite of legislation facilitating the exchange of previously internalized information for crime and disorder reduction purposes (Section 115, Crime and Disorder Act, © 2006 by Taylor & Francis Group, LLC 400 GIS for Sustainable Development 1998), some agencies may be reluctant to provide personalized information for multiagency use The HASCADE model attempts to... to deter partner involvement, as not all partners are able to produce information in the detail required (for example, highresolution full-address information), nor would their service be motivated to respond by purely situational means HASCADE attempts to reduce the risk of such an imbalanced approach to community safety strategy by informing decisions based upon various vulnerabilities, for example,... HART Publishing, Oxford, UK, 2000 18 Home-Office, Review of Crime Statistics: a Discussion Document 2000, http://www homeoffice.gov.uk/rds/pdfs04/review.pdf 19 Ordnance-Survey, Code-Point® Data Southampton, UK, 2001 20 Galton, F., Co-relations and their measurement, chiefly from anthropometric data Proc R Soc Lond., 45, 135–145, 1888 21 Crime-Reduction-Website, Crime Reduction Toolkits, Partnership Working,... considered to be the energy of the future 417 © 2006 by Taylor & Francis Group, LLC 418 GIS for Sustainable Development The duration of solar radiation and the energy that reaches the ground thus represent a very important piece of spatial information Different aspects of solar radiation research are significant for meteorologists, foresters, agronomists, geographers, and other practitioners Solar radiation is... culture (vineyard and olive-grove) for their relevance in typical local productions and for their potential recovery of environmentally sustainable productive traditions With this method a second thematic layer has been created for the six Eolian Islands in terms of macro-localization suitability of new water infrastructures on the basis of results of the Agricultural and Forest Study (Figure 23.5)... through a buffering analysis of the existing buildings) gives much important information both for a closer localization analysis and for compensative actions to provide for impact minimization The procedure is also a first step of a decision-making support tool based on a more formalized multicriteria suitability analysis [9] An ex-ante assessment of environmental compatibility during first stages of public . Group, LLC 390 GIS for Sustainable Development protocols were enforced, whereby no information was to be published (cartographic or text based) beyond the confines of the CDRP without formal approval. analysis is likely to deter partner involvement, as not all partners are able to produce information in the detail required (for example, high- resolution full-address information), nor would their. Group, LLC 386 GIS for Sustainable Development was placed on these CDRPs, particularly the local authority and police, to work in tandem to develop, publish, and implement three-year strategies