Bike Theft in Toronto - Insights from Toronto Police’s Open Data Program
Published: 28-01-2025
Returning to campus after the winter break in my first year, I was ready to lock in and tackle my second semester with an optimistic mindset. You can then imagine the blow of dread as I went to the bike storage of my residence to see no trace of the bike my mother gifted me in high school.
I had great memories of that black and blue Giant Escape. I would ride with my best friend along the lakeshore from Port Credit to Downtown Toronto, exploring the city for the first time without the burden of gas prices, car insurance, or parking fees. Especially during my first year at U of T, riding my bike at night while listening to Gorillaz or The Pharcyde was the ultimate outlet for stress about upcoming assessments.
So after I realized my bike was gone, I decided to call Campus Police, figuring that filing a report would be better than doing nothing. And who knows? Maybe the bike would turn up one day. Two officers came to document the theft, taking down information about the colour, make/model, and where it was stolen from. Nearing the end of the encounter, they suddenly received a call from dispatch and ran off. Turns out there was a bomb threat at a nearby building, but it was later revealed that it was a prop explosive.
I never heard back about my bike. I imagine it’s still out there somewhere. It was this train of thought that led me to follow up on my bike after many years, and seeing if anything came from my police report. To my dismay, the Canadian Police Information Centre’s database for stolen bicycles returned no result, suggesting that my report was not submitted to this database.
Moreover, it turns out my bike theft was never recorded by the TPS, according to their Bicycle Thefts Open Data.
I was visiting my sister in New York City, and it was during one of our countless yap sessions on the subway or bus when she suggested that I look into municipal open data sets. She introduced me to NYC Open Data, an initiative that promotes municipal accountability via a platform accessible by anyone. The amount of data is incredible. Some personal favourites of mine are the 2018 Central Park Squirrel Census, Parking Violations Issued, and Department of Buildings Permit Issuance(s). Adrienne Schmoeker, Senior PM for Open Innovation Initiatives at the Mayor’s Office notes that the city’s snowplows even have sensors that feed data to the platform about what streets have been serviced.
You can imagine the power of this data in the hands of concerned citizens and online researchers who, for instance, can investigate potential connections between those who receive building permits and those who issue them. However, this provides an opportune moment to also discuss the ability for these datasets to be abused. For instance, the previously mentioned Parking Violations dataset includes the Plate ID, car make, and street code of the violation. A nefarious actor could plot out locations where someone’s car was ticketed with only a license plate. Thinking beyond individual actors, the discussion of state data collection is ongoing and vibrant; especially in the city of New York. While expansive surveillance networks such as those enabled by the NYPD’s Argus cameras are restricted to internal access, the shoddy track record of government agencies in securing their databases suggests that compromising citizen data is potentially within the reach of criminal actors and nation-state adversaries.
Returning to Toronto, I was inspired to explore my own city’s open data initiatives. As a result, I created a dashboard that attempts to deliver a broad overview of some major trends in Toronto Bike Thefts. While the TPS presents their own PowerBI dashboard with comprehensive filtering features, I intend to explore some shortcomings of their data visualization methods rather than provide a substitutive dashboard.
Firstly, my dashboard includes collected data about the types of bikes stolen. This helps answer the questions: “What kinds of bikes are stolen?” (An especially relevant question considering the increasing popularity of expensive e-bikes), and “What is the cost of bike theft on Torontonians as a whole?”
Secondly, my dashboard opts to present stolen bike data according to the date of the theft, rather than the reporting date approach of the TPS. This reporting date approach introduces factors that warp the reality of bike thefts in Toronto, such as the variance in reporting time between individuals. This also brings into question the amount of bike theft data that is left uncaptured. My own experience presented a scenario where the citizen followed the expected protocol, yet the data was unrecorded. There is undoubtedly a greater number of citizens whose bikes are stolen and for some reason do not attempt to report the crime. These reasons could include a mistrust of law enforcement and personal time constraints.
Thirdly, my dashboard presents a heatmap. This medium of data visualization is far more accessible to a broader audience of citizens, accounting for our city’s diversity in education levels and language proficiency. Users can easily filter by year, and zoom into their neighbourhood, gaining insights into the most popular intersections for bike theft. In the pursuit of digital accessibility, I have also ensured that the dashboard and this entire website are optimized for both desktop and mobile use.
Next, I will discuss some interesting findings from my research and potential implications.
The costs of bike theft - a shocking revelation! The total cost of reported bike thefts in 2024 was approximately $2,629,430 CAD. Additionally, the average reported cost of a stolen bike in 2024 was approximately $1037 CAD. Even considering a potential inflation of these prices from people using retail rather than market value, this number puts into perspective the massive costs on Torontonians from these crimes. These figures present the financial cost of bike theft, but fail to capture the broader effects of bike crime. I speculate that individuals who live in areas with a high frequency of bike theft have reduced faith in law enforcement to protect their property. These feelings may also contribute to a broader sense of personal insecurity in response to the visible limitations of law enforcement, despite the fact that bike-thefts are largely a non-violent crime.
As suggested in the beginning of this article, the occurrence of e-bike theft is increasing steadily. From 2017 to 2023, we see an increase of annual e-bike thefts from 65 to 263. This is expected given the popularity of e-bikes for personal transport and intercity deliveries. Considering the fact that they often cost 2-3x more than a non electric bike, I hypothesize that a significant factor pushing average stolen bike cost in the past decade has been the increased popularity of e-bikes. Somewhat ironically, the falling price of e-bikes has also almost certainly increased the popularity of their use. It should be noted that 2024 witnessed a significant decrease in reported e-bike thefts, with only 102 reported, which is a -61.22% rate of change.
Bike theft occurrences have been steadily decreasing. From 2020 to 2024, the rate of change is approximately -31.51%. This could be due to many factors such as the rapidly expanding bike share infrastructure and ridership, a preference of taking transit rather than cycling, and a growing awareness of theft prevention. The impact of bike sharing specifically, cannot be understated. Considering the fact that an overwhelming majority of Torontonians are riding bikes in the sub-$500 CAD price range, these bikes provide a convenient substitute for riders of personal bikes while seemingly eliminating concerns of theft with their docking stations.
Bike theft data over the years reveals a consistent trend correlation with seasons and time of day. Specifically, bike thefts tend to begin increasing in April, where it often reaches 200 thefts in the month, in contrast to the 50-100 theft lows of the winter months. Summer months - July and August are peak months for bike theft, often entering the 300-400 theft range. September generally witnesses a significant decrease in bike thefts as the fall/winter season approaches. For time-of-day, bike thefts have the lowest occurrence at around 4 & 5 AM, with a consistent peak at roughly 6 PM. One staggering figure puts these trends into perspective: since data collection started, thefts occur at 6 PM at roughly 5x the rate they occur at 4 & 5 AM.
I intend to return to this project in the future and expand on the following areas:
As mentioned previously, expanding neighbourhood and micro-level insights to empower everyday citizens. Also ensuring site-wide accessibility to screen-readers.
Explore how this data can be cross-linked with demographics, income, transportation accessibility, etc.
Explore how this data can be cross-linked with the rapidly expanding network of bike sharing in Toronto.
Include the complete dataset from 2025. (Updated 01-29-2025 to include the complete 2024 dataset)