Exploring Methods in Crime and Safety Analysis

Abstract

Understanding the spatial distribution of crime patterns or poor safety perceptions equips us with valuable insights to guide resource allocation. In this special issue, we present examples of articles in environmental criminology, focusing on the use of spatial and temporal-based methods for crime and safety analysis. The contributions are from Sweden, Canada, New Zealand, and Pakistan, showcasing a diverse array of data and methods. Almost all articles are from sessions presented in the International Symposium in “Environmental Criminology and Crime Analysis” ECCA 2023, organized by the guest editor's university, and in “the Stockholm Criminology Symposium” that took place in Stockholm in June 2023. This introduction finishes with a summary of the articles of this collection.

Environmental criminology is a field that has a direct impact on the way communities and places are shaped. Defined nearly five decades ago by Brantingham and Brantingham (1981, p. 2), environmental criminology is the field that “argues that criminal events must be understood as confluences of offenders, victims or criminal targets, and laws in specific settings at particular times and places.” Methods dealing with space and time are essential for environmental criminology because they enhance the researchers’ ability to understand, predict, and prevent criminal behavior, enabling proactive law enforcement strategies. Moreover, they serve as invaluable tools for urban planners and policymakers in making informed decisions that contribute to safer and more secure communities.

In this special issue, we present articles that exemplify the research in environmental criminology, employing spatial and temporal-based methods for crime and safety analysis. The word “spatial” is loosely defined to incorporate methods dealing with features of the physical and social environment of a facility but may also include large units of territories such as administrative zones and municipalities up to continuous spaces in satellite imagery. The temporal scale also varies, from minutes in an individual's daily routine activity to a yearly scale in international crime surveys. This collection attempts to discuss methodological challenges when using these approaches, including issues of data availability, resolution limitations, and potential privacy concerns, to name a few.

The articles in this special issue are inspired by various environmental criminological theories. Several articles refer to social disorganization theory, which traditionally has sought to explain crime concentrations in neighborhoods by suggesting that the structural and social characteristics of these areas promote criminal activity (Bursik Jr & Grasmick, 2002; Shaw & McKay, 1942). Social trust is essential for fostering informal social control in neighborhoods, thereby increasing the willingness to intervene for the common good (Sampson, 2006). Research reported in this issue also relies heavily on crime opportunity theories, which focus on the specific settings where crimes are committed. These theories emphasize the significance of environmental factors that precede human behavior and actions, including crime. Crime Prevention Through Environmental Design (CPTED) and situational crime prevention theory are two approaches in criminology focused on reducing crime by altering environmental and situational factors (Clarke, 1995; Crowe & Fennelly, 2013).

Almost all articles are from sessions presented at the 2023 International Symposium in “Environmental Criminology and Crime Analysis,” organized by the Guest Editor's University, and the “Stockholm Criminology Symposium” held in June 2023. Articles are written by environmental criminologists from different disciplinary traditions (sociology, criminology, economy, geography, architecture, and planning) and cover a wide range of international contributions from Sweden, Canada, New Zealand, and Pakistan, showcasing a diverse array of data and methods. All articles were reviewed using a double-blinded peer-review process; we received eight papers, and five are now published.

Framing the Articles of the Special Issue

In this special issue, we highlight innovative research in environmental criminology using spatial and temporal methods for crime and safety analysis. Using a case in New Zealand as a reference, Curtis-Ham and colleagues used discrete spatial crime location choice models to test roles of the reliability and relevance of offenders’ knowledge of locations in their crime location choices. In this first paper, authors analyzed offenders’ pre-offense activity locations recorded in police data (home addresses, family members’ home addresses, work, school, and locations of prior offenses, victimizations, noncrime incidents, and other police contacts) for 17,054 residential burglaries, 10,353 nonresidential burglaries, 1,977 commercial robberies, 4,315 personal robberies, and 4,421 extra-familial sex offenses. Their findings confirm the importance of including both reliability and relevance factors when modeling or predicting offenders’ crime location choices from their activity locations.

Also, looking at residential burglary, Emeno and colleagues test the boost account for repeat and near repeat burglary developed by Johnson and Bowers (2004) in Edmonton, Canada. The theory suggests that these burglaries occur because the same offender returns to burglarize a dwelling they have successfully targeted before or a location near the previously victimized site. In this second paper of the special issue, the authors found a significant repeat and near repeat space-time pattern in the burglary data, meaning that the same residence was at a significantly increased risk of being victimized again a period of seven days following the initial burglary. They also found that pairs of burglaries occurring within 800 meters and 14 days of one another had the highest chances of being committed by the same offender(s).

Using multilevel negative binomial regression, Doyle and Gerell (2024) assessed the predictive accuracy of several characteristics of the crime (using prior crime, place attributes, ambient population, community structural, and social characteristics) to forecast different violent and property crimes. The third paper they found that areas with a history of crime and those with characteristics conducive to crime ran a higher risk for future crime. When considering geographical areas, combining crime history with place and neighborhood characteristics achieves similar accuracy to using crime history alone for most crime types and hotspot cutoffs.

The fourth article examines the impact of building design on perceived safety at a large shopping mall in Karachi, Pakistan. Using Crime Prevention Through Environmental Design (CPTED) principles and an online survey, Iqbal and Humaira evaluate how design, familiarity, and past victimization impact visitors’ safety perceptions. They also critically reflect on the advantages and disadvantages of online surveys and CPTED principles to assess safety in shopping centers.

Researchers and planners often rely on fieldwork inspections, surveys, and interviews to collect detailed spatial data on crime environments. Still, these methods can be time-consuming, labor-intensive, costly, and dependent on timing. This is especially challenging in resource-poor urban areas where conventional data is scarce, complicating the study of criminogenic conditions. As an alternative, remote sensing data, such as satellite images and drone photographs, offers a noncontact method of data collection. In the final article of the special issue, Ceccato and Ioannidis report on the use of remote sensing for analyzing criminogenic conditions in urban environments through a systematic review of English-language literature from 2003 to mid-2023, examining 36 articles. They find that remote sensing data, often available for free, can aid crime analysis where conventional data is lacking. At the same time, future research should focus on developing and accessing new tools for urban analysis, such as image feature extraction for streets, property boundaries, housing density, vegetation characteristics, and luminosity levels. Machine learning algorithms can predict crime probabilities using historical data and remote sensing information, while object detection algorithms can identify environmental features associated with crime and accidents. Future research should also deal with concerns about privacy infringement and validation of methods.

All these papers share several key commonalities that highlight their contributions to environmental criminology. They each focus on how environmental factors influence crime and safety, employing spatial and/or temporal-based methods to analyze crime patterns and predict future incidents with studies by Curtis-Ham et al. and Doyle and Gerell. The studies are data-driven, relying on extensive datasets, including police records, surveys, and remote sensing data, to draw conclusions. Additionally, the impact of design and the environment is a recurrent theme, as seen in the works of Emeno et al., Iqbal and Humaira, and Ceccato and Ioannidis, which explore how physical and social environments, including building design and urban planning, affect crime levels and perceptions of safety. These papers also address methodological challenges, such as data availability and accuracy, and propose the use of new data sources and machine learning algorithms to enhance crime analysis and prevention strategies. Finally, the diverse findings and approaches outlined in these articles can be directly linked to several of the United Nations Sustainable Development Goals (SDGs) (UN, 2015), demonstrating the role that environmental criminology and situational crime prevention play in achieving broader global objectives in promoting sustainable cities and communities.