Crime Metrics for the People: Measuring Crime for Safer CommunitiesBy Madison Charman and Samantha Scott, Barrie Police ServiceThe way we measure crime in Canada is flawed. Traditional police metrics focus on counting criminal and non-criminal occurrences, equating volume with public safety. This approach provides a one-dimensional view of a multidimensional issue. Crime volume measurements are susceptible to both internal and external influences, leading to skewed data. Sherman (2020) notes that police efforts concentrated on a specific concern, or private resources dedicated to loss prevention can artificially inflate crime numbers.This method can unintentionally penalize good police work. For example, if the Barrie Police Service (BPS) forms a task force to tackle drug and weapon trafficking, resulting in multiple charges through proactive efforts, the data will reflect an increase in crime volume. This does not mean more crime is occurring; it simply indicates that the police have uncovered more offences due to targeted initiatives. Unfortunately, this spike in charges can create a perception that crime is on the rise, leading people to say, “crime is out of control,” when in reality, it is more visible and detected due to better policing.A similar issue arises when a large department store hires additional loss prevention officers. With more resources, they detect and report more incidents of shoplifting, which leads to a sharp increase in crime statistics. Again, this creates the false impression that crime is worsening, when the reality is that more resources were put into detecting it. In both cases, the perception of safety declines, even though underlying crime rates may not have changed at all.A NOTE ON LEGITIMACYOur data is only as reliable as the community’s trust in law enforcement to report their experiences. When the community has confidence in law enforcement, individuals are more likely to report crimes, leading to more comprehensive and reliable data (Mazerolle, 2014; Bradford, 2012; Tyler, 1990). Conversely, when trust is lacking, many crimes, particularly those involving vulnerable or marginalized groups, go unreported, creating a significant gap in official crime data and distorting the true picture of public safety. This “dark figure of unreported crime,” as coined by Skogan (1977), can misinform resource allocation and policy decisions. Trust also affects the community’s perception of public safety – when people believe crime data better reflects reality, it strengthens police legitimacy. This is an inherent limitation of police data and any analysis based on it. Therefore, when we discuss the harm index, we will refer to it as the police-reported crime severity index (PRCSI).A NEW MEASUREMENT: HARMHarm-focused policing is rooted in the idea that not all crimes have equal impacts on victims or communities. The most significant issue with current policing measurements is that every incident is treated equally – counted as a single occurrence regardless of its context or severity. Both minor and serious crimes are treated equally in the data, which doesn’t reflect their differences in severity or the resources needed to respond. For instance, having a credit card stolen versus being a victim of assault with bodily harm are vastly different experiences, yet both count as just one crime statistic. This fails to emphasize their differences in severity and the resources needed for response. While this measure provides a narrative about crime occurrence, it is unhelpful for understanding how serious the crimes are.Harm Relevance to DeploymentPolice agencies use crime data to make decisions about resource allocation. If crime is measured solely by volume, resources might be disproportionately directed toward high-volume, low-severity crimes, leaving more severe or harmful crimes under-addressed. Not everything that occurs within a criminal environment should be assessed under the same lens because they do not pose the same level of threat or require the same resources.Map A displays the total count of CAD (computer-aided dispatch) events across the city, including calls for service, criminal and non-criminal incidents. While it highlights areas with a high volume of incidents, it doesn’t provide much insight into the actual nature or severity of these events. Map B represents the same area, but it focuses on events with UCR codes that carry a harm score. This map reflects the count of incidents deemed “harmful,” weighted by severity, offering a different perspective. Map C also covers the same area, but it specifically shows harm levels associated with each incident, using a weighted harm score to provide an even more focused view on threat levels. This maps the weight of harm.Map A (traditional police measurement) and Map C (harm-focused measurement) tell two very different stories about where resources should be deployed and what kind of resources are needed in those areas. Map A, showing social disorder hotspots driven by volume, may suggest the need for community engagement teams, special constables or partnerships with local agencies. In contrast, Map C identifies high-harm areas, which would likely require a more robust police presence, potentially with specialized teams like CAP units or enhanced surveillance. This comparison underscores the importance of distinguishing between event volume and actual harm when determining resource allocation.WHAT IS A HARM SCORE?A harm score is a weight assigned to a crime based on its severity, interpreted as the average days spent in jail for committing that crime (Sherman, 2016). For example, the harm score for assault causing bodily harm is 501.06, meaning that on average, those convicted serve 501.06 days in jail. For theft, the harm score is 29.34, indicating that those convicted typically serve 29.34 days. The more serious the crime, the higher its weight.NOT ALL CRIMINAL EVENTS REQUIRE THE SAME LEVEL OF RESOURCE ALLOCATION, AND GROUPING HARMFUL INCIDENTS INTO A SINGLE CATEGORY CAN OVERSIMPLIFY THE ISSUE AND DIMINISH THE GOOD WORK DONE BY RESOURCES.The Barrie Police Service uses the Crime Severity Index (CSI) as the source for the weights. As noted above, this is limited to police-reported crime, leading to the Police Reported Crime Severity Index language (PRCSI) (Babyak et al., 2009). The PRCSI measures the seriousness of crimes reported to police, giving more weight to serious offences, like violent crimes, compared to less serious ones, like minor thefts (Wallace, 2009). The CSI was created in 2009 with personnel from Statistics Canada, the Canadian Association of Chiefs of Police (CACP), federal parliamentarians and academics. Statistics Canada updates the values every five years based on actual custodial sentences handed down across the country, calculating the average number of days served in incarceration for each offence. This is then multiplied by the incarceration rate for that offence.The key for police services is that we cannot influence the Crime Severity Index (CSI). This objectivity adds credibility and legitimacy to its use. Harm weights in the CSI are determined by the judiciary, whose roles are assigned by democratically elected governments, making the CSI a reflection of societal values. These weights are updated every five years to account for changes in judicial sentencing patterns and ensure a sufficiently large sample of sentences is collected, producing a more accurate national average. The CSI focuses solely on custodial sentences, excluding non-custodial penalties like suspended sentences (Wallace et al., 2009; Babyak et al., 2009).Types of HarmHarm is not a one-size-fits-all approach. Based on research and literature, harm should be broken down further. Not all criminal events require the same level of resource allocation, and grouping harmful incidents into a single category can oversimplify the issue and diminish the good work done by resources. Research suggests further breaking down harm indices into proactive, commercial and community-reported harm (Sherman, 2020).Community-Reported HarmCommunity-reported harm includes offences against individuals or property reported to the police by the community, independent of law enforcement or private influence. Examples encompass serious violent crimes like murder, sexual assault and robbery, along with impaired driving causing death, fraud, theft and breakand-enter incidents. These offences prompt reactive police responses and indicate genuine community safety concerns.Proactively Detected HarmProactive harm consists of offences typically identified through proactive police initiatives, including drug possession, trafficking and weapon-related charges. These crimes are often not visible to the community and are less frequently reported by individuals. This type of harm is sensitive and can be skewed by police resource allocation. When law enforcement increases resources dedicated to targeting these offences, detection rates typically rise.Commercial HarmCommercial harm refers to crimes identified and influenced by private resources, with shoplifting being the primary example. The use of advanced technology and resources for detecting commercial crime can lead to an increase in reported incidents. However, this rise does not necessarily reflect a true increase in crime rates; rather, it highlights the extent of investment in detection efforts.A Note on HarmThrough leveraging this method of harm measurement, we realize the context of certain offences can determine where it is placed between the three categories. For example, weapons offences. Offences involving weapons can be parsed out between community-reported harm and proactive harm. The classification of the specific offence – whether as community-reported or proactively detected – depends on whether the weapon was used in the commission of the offence or simply found in someone’s possession. In the first example, the police responded reactively after the weapon was used in an offence. In contrast, the second example demonstrates proactive policing, where officers detected a weapon in someone’s possession before any offence occurred.Part of better understanding the complexity of crime in our communities is seeking to be more accurate in our recording of crime. It is important to calculate the crime harm that occurs within the year that it occurred rather than what year it was reported, in those situations when they are not one and the same. Recording of historical offences in the year they are reported distorts the crime harm in the year reported and underreports the harm in the year it occurred. (Sherman et al., 2020)Additionally, we acknowledge that harm is complex, and all criminal offences can cause varying degrees of harm. While it’s impossible to capture every nuance of victim experiences, this approach is an improvement over relying solely on crime volume. For example, the theft of a family heirloom and the theft of a lawnmower may both be recorded as a single crime, yet their impacts differ significantly. The PRCSI, while imperfect, serves as a valuable tool for law enforcement, offering critical insights into our crime landscapes. When interpreted in context, it enhances our understanding of the real-world impacts of crime in Canada.CONCLUSIONThe traditional metrics of crime measurement in Canada fall short of capturing the complexities of public safety and community well-being. By shifting our focus from mere volume to the additional consideration of a focused approach, we can better understand the nuances of crime and its impacts on individuals and communities. The Police-Reported Crime Severity Index (PRCSI) provides a more meaningful framework that recognizes the differing severities of offences, and the resources needed to address them. By categorizing harm into community-reported, proactive and commercial harm, we can allocate resources more effectively, fostering a proactive police response that prioritizes genuine public safety concerns.Madison Charman works as an Organizational Researcher at the Barrie Police Service. With a BA, MA and PhD (in progress) in criminology, she blends her academic expertise with a firsthand understanding of police operations. Madison has been working with the Barrie Police Service in applied research since 2020. She can be reached at mcharman@barriepolice.ca.Samantha Scott is a Tactical Operations Crime Analyst at the Barrie Police Service’s EvidenceBased Research and Innovation Unit, where she has significantly contributed to enhancing evidence-based policing practices. She can be reached at sscott@barriepolice.ca.ReferencesBabyak, C., Alavi, A., Collins, K., Halladay, A., & Tapper, D. (2009). The methodology of the police-reported crime severity index. Ottawa: Statistics Canada.Sherman, L., Neyroud, P. W., & Neyroud, E. (2016). The Cambridge crime harm index: Measuring total harm from crime based on sentencing guidelines. Policing: A Journal of Policy and Practice, 10(3), 171-183.Sherman, L. W., & Cambridge University associates. (2020). How to count crime: the Cambridge harm index consensus. Cambridge Journal of Evidence-Based Policing, 4(1), 1-14.Wallace, M. (2009). Measuring crime in Canada: Introducing the crime severity index and improvements to the Uniform Crime Reporting Survey. ProQuest.Weinborn, C., Ariel, B., Sherman, L. W., & O’Dwyer, E. (2017). Hotspots vs. harmspots: Shifting the focus from counts to harm in the criminology of place. Applied Geography, 86, 226-244.