Crime Volume in Toronto
Analysis performed by Inna Semenykhina
Analysis performed by Inna Semenykhina
Audience: Government, Law Enforcement, Businesses and Local Communities
Key goal: To provide a deeper understanding of crime trends over time and a reliable forecast for implementing appropriate action.
Value to stakeholders:
Government and Public Safety: evaluate policy success, plan budgets, and develop safety projects.
Business and Investment: identify the safest areas for development and investment.
Communities and Non-Profits: organize localized safety plans and resource requests.
The charts below reveal the top 10 neighbourhoods with the highest total crime volume.
When evaluating areas for resource allocation, it is crucial to distinguish between raw crime volume and the crime rate per resident. Based on the raw total crime count, neighbourhoods like Mimico-Queensway, West Humber-Clairville and Moss Park rank highest due to the sheer number of incidents. However, when crime is measured per 100,000 residents — providing a clearer picture of individual risk — the highest risk areas shift significantly. The Yonge-Bay Corridor exhibits the highest crime rate by a significant margin, making it the area where an average resident faces the highest probability of being impacted by crime. Therefore, while total volume informs policing deployment to busy hubs, the crime rate should guide community safety programs and long-term planning focused on reducing risk for the local population.
The chart below shows the crime trend by type (2014–2024 years), highlighting the three neighbourhoods with the highest crime rate for each crime type.
Analysis of the historical data reveals that crime volume is not static, but rather shifts significantly in composition over time. The most notable trends are the consistent dominance of Assault and Theft from Motor Vehicle as the largest categories, forming the base volume of crime in most high-volume neighbourhoods. The period between 2021 and 2024 shows a sharp, concerning rise in Auto Theft, becoming a major contributor to the overall volume. In contrast, crime types like Robbery and Break and Enter have remained relatively stable or shown more minor fluctuations. This mixed trend indicates that while overall volume may appear stable, resources must be dynamically shifted to combat the specific crime surges, particularly the escalating issue of assault and vehicle theft in recent years.
City-Wide Planning: The model's forecast for total aggregate crime across the city is highly reliable, with an average prediction error of just 14.7%. This forecast is recommended as the foundation for annual budget and city-wide policy decisions.
Targeted Allocation: The model's forecast for individual neighbourhood crime volume has an average prediction error of 18.1%. This data provides a strong directional guide for deploying resources to specific high-risk or high-growth areas.
Visualization 1: Crime Volume per Neighbourhood in Toronto with Forecast for 2025
Insight: This chart shows the ten-year trend (2014-2024) for any selected area, with the 2025 forecast point clearly highlighted.
Action: Immediately identify neighbourhoods where the line shows a steep upward trend, signalling a growing and urgent need for preventive resources.
Visualization 2: Crime Volume for Each Crime Type in Toronto with Forecast for 2025
Insight: This metric view provides the most actionable data by showing the trend for each crime type over time for comparison including forecast for 2025 year.
Action: Initiate a strategic review of current crime reduction programs targeted at the top three crime types identified in the 2025 forecast. Policy adjustments are required for any crime type where the forecast trend indicates that the current strategy is failing to mitigate year-over-year growth.
Visualization 3: Crime Volume Trend Over Time in Toronto (Top 3 Neighbourhoods)
Insight: The visualization reveals long-term trends (2014–2024 years) and the proportional contribution of the top 3 neighbourhoods to the total crime volume for eight distinct crime types.
Action: Develop a tiered investment strategy that prioritizes preventative and enforcement measures in areas identified as both dominant crime contributors and those exhibiting adverse emerging trends based on the crime volume data and 2025 forecast. Focus initial deployment efforts on mitigating high-severity crimes.
Additional recommendations:
Prioritize the Aggregate Forecast: Utilize the highly accurate Total City Crime Forecast (14.7% error) as the non-negotiable basis for establishing annual police budgets and assessing the effectiveness of city-wide policies.
Targeted Allocation of Resources: Use the Per-Neighbourhood Forecast (18.1% error) to guide strategic resource allocation. Deploy additional patrols or launch community engagement initiatives in areas where crime is specifically projected to rise.
Future Data Maturity (ARIMAX): The next step for improved accuracy is to move beyond relying solely on historical crime counts. We recommend integrating external factors, such as socio-economic indicators (e.g., unemployment, housing instability) or police staffing levels, into the model to achieve a deeper understanding of the causes and drivers of crime trends.