Thames Gateway Crime and Thames Gateway Design Project Crime and Design Project Mapping neighbourhood vulnerability to street crime 3rd National Crime Mapping andConference burglary Context • Thames Gateway represents the largest and most ambitious regeneration project in Europe. • 91,000 homes and 150,000 jobs • unique city-building opportunity. • exemplar of the standards and principles of sustainability, cohesion and liveability expected by the government across the UK • high levels of crime, or simply the fear of crime, will make the benefits of regeneration inaccessible or unsustainable. • principles of successful community and neighbourhood development should be built into the regeneration process from the very start. • once new communities are established it will be difficult to rectify problems and try to address community and design issues later. The Crime and Design Project 3 strands building towards development policy and practice 1. 2. 3. Theme guide – a regeneration framework to inform development of project briefs, project appraisal and evaluation, decision making, for the development of housing, open space, town centres etc Concordat – a base-level agreement between built environment practitioners and the police to work closer together and in the early stages of a development proposal Mapping project – analysing crime data at a neighbourhood level to ascertain what environmental and socio-economic attributes count as factors in the vulnerability of neighbourhoods to crime Critical Questions: • What makes a “good” neighbourhood? • Why are some neighbourhoods more vulnerable to crime than others? • Are there environmental and social factors that contribute to a neighbourhood’s vulnerability? • Can we isolate those factors which contribute more than others? • How perceptive are environmental variables? • Can we develop rules or patterns that govern vulnerability to certain types of crime? • Can these rules in turn inform planning and design solutions to effectively remove them? • What will be the impact of this on the ability to develop sustainable communities? Mapping neighbourhood vulnerability Considers 3 groups of ideas underpinning the propensity for crime in neighbourhoods and uses them as a starting point to test the following statements: 1. Areas will not regenerate if people are not attracted to them. Neighbourhood satisfaction is judged by resident’s perception of cleanliness, vandalism and crime levels. Has an effect on the attractiveness of a neighbourhood to potential new residents. (broken windows) 2. Lively environments are safer environments. Environments that have been designed with safety and crime reduction in mind are more successful places because they build in overt control and management of spaces to a distinct group of users. (defensible space) 3. Cohesive communities are self-policing. Developments that have considered the impacts of design on the people that use them and have catered for their needs with correct level of facilities will lead to more cohesive communities. (social efficacy) Thames Gateway London Deprivation Thames Gateway London boroughs Dependent variables • analysed 2 dependent crime variables across the Thames Gateway : i. annual rate of street crime per kilometre of street ii. annual rate of domestic burglary per thousand households. • *Note: Street crime does not include anti-social behaviour • Presented at Census Output Area (COA) level • Separate from IMD crime dimension which looks at all types of crime Independent variables • 30 independent variables selected • identified in literature as being essential for developing cohesive communities • suggests that physical planning, management of the fabric and social composition all play a part? • 14 variables from the census • 11 represent distances to a range of facilities • calculated as shortest distance from each postcode Environmental Variable Description Source Vacancy Proportion vacant properties ONS Deliveries Non-domestic deliveries as a proportion of total deliveries OS Dentist average weighted distance to a dental practice TGLP GP average weighted distance to a GP practice TGLP Pharmacy average weighted distance to a pharmacy TGLP Police average weighted distance to a police station MPS Rail average weighted distance to a rail/underground station TFL Post Office average weighted distance to a post office Royal Mail Town Centre average weighted distance to a town centre OPDM Newsagent average weighted distance to a newsagents or store Yell.com School average weighted distance to a primary school boroughs Open space average weighted distance to open space Green Grid + Bartholemew Bus average weighted distance to a bus route TFL Density Residential density per hectare from buffered postcodes ONS + OS Street length Average length of street section Bartholemew Street density Street density metres per hectare Bartholemew + OS Socio-Economic Variable Description Source Ethnic minority Proportion ethnic minority ONS Female lone parents Female lone-parents as proportion of households with dependent children ONS Foreign born Proportion foreign born ONS Social grade Proportion social grade C2 to E ONS Young people Proportion population aged 15-24 ONS Low qualifications Proportion population with qualifications < Level 3 ONS Low qualified young people Young people 18-24 with qualifications < Level 3 as proportion of age group ONS Religious minority Proportion of population religious minority ONS Unemployed Proportion of economically active that unemployed ONS Carers Proportion of population providing unpaid care ONS Turnover Residents not at this address in previous year as proportion of total population ONS Migration Net migration as a proportion of total population ONS IMD score Interpolated IMD 2004 from Super Output Area Centroids ODPM Renting Proportion rented accommodation ONS Analysis: Correlations • • • • Spearman rank correlations: 1. street crime: some strong positive and negative 2. burglary: weaker correlations and two variables (vacant dwellings and open space) are not significant. correlations >=+0.2 or <= -0.2 qualified for a factor analysis exploratory factor analysis for each crime type to weed out variables with low factor scores 5 factors/dimensions were identified for each crime type, broadly similar but ordered differently Analysis: Regression Exercise in isolating the most significant variables using increasingly detailed regressions until the most significant variables are isolated: 1. Create indices to represent the 5 dimensions 2. Comprise the factors with the highest scores 3. Regression tests the correlation between indices and the crime variables 4. Results are run through an induction programme to generate decision trees for each crime type 5. first time that these type of decision trees have been derived for crime types at the neighbourhood level Analysis: Decision trees • • • • • • Measuring characteristics of vulnerability Depending on degree of significance, the indices generate different sets of relationships Gives rise to a probability for crime according to branches followed splits in the tree reflect the multiple correlation scores Helps us to identify most significant LOCAL factor(s) Can then work towards designing/managing them out Example: Street crime decision tree Some findings: • Closer a neighbourhood is to a facility/service the greater risk of burglary • Shorter the street the lesser the risk of burglary • Greater the street density the greater risk of street crime Supports defensible space theories However!!!!!! • Physical factors explain only <50% of the story • Reinforces idea that environmental design alone can only do so much • has to be supported by people management Street Crime 2002: Neighbourhoods rates at Census Output Area level Household Burglary 2002: Neighbourhoods rates at Census Output Area level Distances to community facilities Street crime • Immigrant/minority groups is the most important dimension for street crime • followed by deprivation/exclusion and low order centrality. • *Note: not saying that these are the groups engaged in crime • this dimension is a proxy for: – there is segregation between minority groups within a neighbourhood? – these areas have more restricted life chances? – lack of assimilation into mainstream society? – difficulty in engaging with agencies? Implications: • Reinforces literature that states community efficacy is one of the most important factors in creating sustainable and safer communities • Reinforces validity of government’s programme for social exclusion • Highlights importance of providing adequate access to facilities in the neighbourhood • Supports the idea that crime is a determined by local factors and must be tackled through local interventions • Can use the model to evaluate new masterplans in the Thames Gateway • Help determine the optimum range and siting of facilities to develop neighbourhoods with sufficient strength to reduce vulnerability to crime