On the first Sunday of November, the clocks go back an hour and sunset lands in the middle of the evening commute. Our study of Toronto police data suggests that shift does something measurable: in the three weeks after the clocks go back, home break-ins between 5 and 7pm rise sharply, while break-ins across the rest of the day barely move. The full study, with the hour by hour charts, is on our clocks go back study page.
Every dot is a place scored in the MkaeSense Safety Index for 2025, plotted on a real map of Canada and the United States. Drag to pan, scroll or pinch to zoom, and select any dot to open its full city profile.
If the map does not load, browse every place in the City Safety Rankings. Data: MkaeSense Safety Index, Data Release 2025.1 (Statistics Canada; FBI Crime Data Explorer).
The finding
We compared the three weeks before and the three weeks after every fall clock change from 2014 to 2025, twelve years combined. Between 5 and 7pm, Toronto recorded 225 home break-ins in the before weeks and 395 in the after weeks, a rise of 76 percent. Across all hours of the day, the change over the same weeks was just 5 percent. The increase is not a general autumn rise in break-ins. It is concentrated in the two hours that newly fall dark.
In spring the pattern reverses. When the clocks go forward and the same evening hours gain light, those hours get quieter. The evening peak in Toronto's break-ins follows sunset in both directions, which is what you would expect if darkness itself, rather than the calendar, is doing the work.
How we tested it
Break-ins rise in autumn for other reasons, so a simple before and after comparison would not prove much. The test instead compares the evening hours against control hours in the same weeks, daytime and late evening hours whose light does not change when the clocks move. That cancels out the general seasonal trend. If evenings rise relative to those control hours at exactly the moment sunset moves, the clock change is the difference.
We also ran the same test on a fake change date, four weeks away from any real clock change, where no effect should appear. It did not. And the pattern shows up in almost every individual year of the twelve, not just in one unusual autumn.
Chicago, the US city we ran the same test on, shows why the effect is local rather than universal. Chicago's home burglaries are spread through the day, with the most around 8am and noon, and no early evening peak for the clock change to move. There, the results sit within the range chance would produce. Toronto has a distinct early evening break-in peak, and that peak is what travels with the sunset. The two cities also record incidents differently, so we compare the patterns, not the totals.
Why darkness at 6pm matters
The early evening is when homes are most predictably empty or in transition: people at work late, out for the evening, or not yet home, with curtains open and lights off. In October, a house sitting dark at 6pm reads as occupied people living normally in daylight's last hour. In November, the same house at 6pm has been visibly dark and still for an hour. The data cannot see a burglar's reasoning, but it can see the timing move, and it moves with the light in both directions, every year, in a way the control tests say is not chance.
This year's clock change lands on Sunday, November 1, which means the three week window the study measures starts now. If the twelve year pattern holds, Toronto's early evenings are at their most exposed between now and late November, in the gap between the sunset people are used to and the one they have adjusted to.
What to do this week
- Move light timers an hour earlier. A timer set to switch on at 6pm in October turns on after dark in November, which defeats its purpose.
- Check outdoor lights and cameras at 5pm, not 6pm. The first weeks after the change are when the evening shift happens.
- Lock up before leaving for the evening, not when you get home. Coming home after dark to a house left dark and unlocked is the pattern this data describes.
- Be most careful in the first three weeks. That is the window the study measures, and the window where the before and after difference is concentrated.
Toronto's wider picture, including its Safety Index of 41 and rank of 74 among 125 Canadian police services, is on its Toronto city page. The break-in tracker follows break-ins across Canada and the US down to the neighbourhood level.
Source: Toronto Police Service open data, clock change study published October 31, 2026, part of MkaeSense Data Release 2025.1.