Crime prevention and safety measures in socio-economically vulnerable areas are important incentives for creating a socially sustainable city. The topic has been debated in Sweden, and to address the challenges, a new law in Sweden will come into force on June 1, 2023, giving Swedish municipalities greater responsibility for crime prevention. Previous research indicates that crime prevention and safety promotion can be categorized based on physical and social efforts. These efforts are partly about creating better social cohesion between people and partly about changing the physical environment to make it more difficult for people to commit crimes. The master thesis aimed to study crime prevention and safety promotion measures in two socio-economically vulnerable areas in Sweden to find out how the work can be developed through the theories of social sustainability, social disorganization theory, and collective efficacy. The results indicate that both neighborhoods work with social and physical measures, but that their circumstances are different. This is because crime statistics indicate that Gottsunda has significantly more crime rates than Klockaretorpet. Finally, the results show that it is important to implement both social and physical measures in socio-economically vulnerable areas. The study was conducted through qualitative interviews with planners, police, property owners, and one social sustainability consultant. A document analysis of policy documents has also been conducted. Furthermore, crime statistics from the police have been analyzed to visualize the number of reported crimes in the neighborhoods

Submitted by Fabio Sobreira Campos da Costa (fabio.sobreira@ufpe.br) on 2017-05-08T14:34:09Z No. of bitstreams: 2 license_rdf: 1232 bytes, checksum: 66e71c371cc565284e70f40736c94386 (MD5) Debora_Pereira_thesis.pdf: 4477202 bytes, checksum: 0af71560a8db492ecea51ebedad76427 (MD5) / Made available in DSpace on 2017-05-08T14:34:09Z (GMT). No. of bitstreams: 2 license_rdf: 1232 bytes, checksum: 66e71c371cc565284e70f40736c94386 (MD5) Debora_Pereira_thesis.pdf: 4477202 bytes, checksum: 0af71560a8db492ecea51ebedad76427 (MD5) Previous issue date: 2016-08-01 / FACEPE / CAPES / In Brazil, since 2000, approximately 50,000 people are murdered every year. In a span of 30 years (1980 – 2010), more than 1 million homicides were registered. In 2012, the homicide rate in Brazil was 29 homicides per 100,000 inhabitants. All Brazilian states exceed the threshold of epidemic established by World Health Organization. In this context, the present study has the objective of to investigate homicides in Recife, taking into account temporal, spatial, environmental, and multicriteria analysis. The temporal analysis shows that the difference of homicides between seasons and months is not statistically significant. However, there is a significant increase in homicides during the weekends (42 percent of all homicides) and evenings (62 percent). Moreover, the spatial results show that the spatial patterns are different within the temporal dimensions in many cases. The findings from spatial analysis reveal that homicides are very concentrated in the city of Recife and in a time span of five years (20092013) all the homicides occurred in less than 10 percent of the street segments. In addition, our test showed that the spatial pattern was not stable over the years. However, when we consider the temporal dimensions (as suggested by temporal analysis), the patterns were stable along the years – except for weekdays and night/dawn. Furthermore, through the environmental analysis, we found that inequality, rented houses, and number of residents have a positive relationship with homicide. On the other hand, income, education, public illumination, population density, and street network density have a negative relationship. The findings of these analyses indicate that homicide in Recife can be understood by the perspective of social disorganization theory and routine activity theory. Finally, multicriteria approach was applied to highlight vulnerable areas to homicide in Recife. We considered six variables to evaluate vulnerability and the areas were identified by PROMETHEE II method and local Moran’s I. Other application was made in Boa Viagem neighborhood, so we were able to perform a more detailed analysis. Three different approaches were tested for Boa Viagem and we suggested some actions in order to reduce criminality in long term. / No Brasil, desde 2000, aproximadamente 50,000 foram mortas todos os anos. Em um espaço de 30 anos (1980 – 2000), mais de 1 milhão de homicídios foram registrados. Em 2012, a taxa de homicídio no Brasil era 29 homicídios para cada 100,000 habitantes. Todos os estados brasileiros excedem o limite de epidemia estabelecido pela Organização Mundial de Saúde. Nesse contexto, o presente estudo tem o objetivo de investigar os homicídios em Recife, levando em consideração análises temporal, espacial, ambiental e multicritério. A análise temporal mostra que a diferença de homicídios entre estações do ano e meses não é estatisticamente significativa. Porém, existe um aumento significante de homicídios durante os finais de semana (42 por cento de todos os homicídios) e noites (62 por cento). E ainda, os resultados espaciais mostram que os padrões espaciais são diferentes dento das dimensões temporais em muitos casos. Os achados da análise espacial revelam que homicídios são muito concentrados na cidade do Recife e que em um espaço de tempo de cinco anos (2009-2013) todos os homicídios ocorreram em menos de 10 por cento dos segmentos de rua. E ainda, o teste do padrão dos pontos espaciais mostrou que os padrões espaciais não foram estáveis no decorrer dos anos. Porém, quando se considera das dimensões temporais (como sugerido pela análise temporal), os padrões foram estáveis ao longo dos anos – com exceção de dias de semana e noites/madrugadas. Além disso, através da análise ambiental encontrou-se que desigualdade, casas alugadas e número de residentes têm uma relação positiva com homicídio. Por outro lado, renda, educação, iluminação pública, densidade populacional e densidade da rede de ruas têm uma relação negativa. Os achados dessas análises indicam que os homicídios em Recife podem ser entendidos pela perspectiva da teoria da desorganização social e da teoria das atividades de rotina. Finalmente, abordagem multicritério foi aplicada para destacar áreas vulneráveis aos homicídios em Recife. Considerou-se seis variáveis para avaliar a vulnerabilidade e as áreas foram identificados pelo PROMETHEE II e pelo índice local de Moran. Outra aplicação foi feita no bairro de Boa Viagem e foi possível realizar uma análise mais detalhada. Três diferentes abordagens foram testadas para Boa Viagem e sugeriu-se algumas ações no sentido de reduzir a criminalidade no longo prazo.

The systemic model posits that informal social control directly reduces crime victimization and social networks indirectly reduce crime victimization through informal social control. While empirical testing of the systemic model advanced the theory, important analytical issues remain. First, social networks are inconsistently conceptualized and measured. Second, the conceptual relationship between social networks and informal social control remains unclear. This study addresses these issues by testing an alternative to the systemic model, including new constructs and hypotheses. The goal is to develop better indicators for the model and refine the theory, rethinking and deepening the existing theory about neighborhood effects on crime victimization. The data come from the 2002-2003 Seattle Neighborhoods and Crime Survey (N=2,200). Structural equation modeling (SEM), a multivariate statistical technique, was used to analyze these data. The SEM included five latent constructs (neighboring, neighborhood and non-neighborhood social networks, collective efficacy, and crime victimization) and six social structural variables (racially homogeneous neighborhood, resident tenure, household income, family disruption, male, and non-white ethnicity). One of my 9 hypotheses was supported; the remaining hypotheses were partly supported. The results support my argument that the systemic model is too simplistic, but the relationships among the variables are not exactly as I hypothesized. The results provide insight into the complexities of the systemic model and areas for future research.

The purpose of this study was to explore the impact of social structure and community organization factors on campus crime on California community college campuses. The study utilized social disorganization theory to examine those relationships by aggregating data from the 113 campuses that are required to submit data on an annual basis to state and federal agencies. Crime data from the 2011 Clery report was separated into personal crimes and property crimes as reported by the California community college campuses. Correlation analysis was used for the non-categorical social structure and community organization factors. Those factors which demonstrated a statistical relationship with personal or property crimes were then regressed to further analyze the data. The relationships of categorical social structure and community organization factors with personal and property crimes were studied using analysis of variance. The results demonstrated that most of the social structure and community organization variables did not have a statistically significant relationship with personal or property crimes. However, the percentage of students receiving general financial aid in the form of Pell Grants demonstrated a strong relationship with an increase in both personal and property crimes. Additionally, an increase in the ratio of part-time to full-time faculty members demonstrated an increase in property crime on campus. The findings indicate that students on campuses with a high percentage of Pell Grant recipients are likely to be impacted by crime in some manner and that campuses with a large percentage of Pell Grant recipients need to consider additional supports and interventions to protect students from crime on campus. The findings also demonstrate that hiring a large percentage of part-time instructors may be counterproductive to preventing campus crime.

Although studies examining juvenile recidivism have focused primarily on violent recidivism, the factors that predict recidivism likely differ by offense type. To examine general, property, and violent recidivism, this study combined individual-level data (i.e., offender and case characteristics) from the Mississippi Youth Court Information Data System (MYCIDS) for the years 2009-2011 and contextual-level data (i.e., county characteristics) from the 2010 U.S. Census and the 2010 Uniform Crime Reports (UCR). Results showed that offender characteristics predicted only general and property recidivism, but case characteristics mattered for all three types (i.e., general, violent, and property recidivism). Contextual characteristics (i.e., the percentage of the population that is male aged 15 to 24) also mattered, but only for property recidivism. These findings have implications for policies and programs related to the treatment of juvenile offenders.