“They are equally safe within statistical margins”

From the context of road safely, this would mean that the roads are not safe, since we still have an overwhelming number of accidents, injuries, and fatalities. But I digress.

To the point of cannabis-related DUI, the link above suggests that impaired driving as a factor to accidents went down 7.7%, while the stats breaking down DUI (the link I posted) still suggests that cannabis-related DUI has gone up by quite a bit.

Would DUI numbers be even lower had cannabis not been legalized? Your guess is as good as mine.

A few rather annoying complications to gathering these stats over the last five years, however, involve COVID and underreporting.

For instance, did officers test people less often because of the risk of close contact during the 2020-2021 pandemic? We know that there were fewer drivers on the road, too, so the numbers can't really be used as a reliable marker one way or the other. Fewer drivers = fewer accidents (but fewer accidents doesn't mean fewer impaired drivers.

For Toronto at least, more people were caught driving under the influence year-over-year during the same time period.

We also know that DUI charges weren't even being pursued due to backlogs.

To me, it seems that there may have been far more people driving under the influence than the stats lead us to believe.

The study I linked also suggests that many samples are obtained quite a while after an accident. So the actual level of drugs in someone's system is also being reported much lower than it would have been at the time of the accident. Couple this with the above point, and things looked much safer than they actually were.