Would intersections be ‘safe’ if pedestrians followed the rules?

A few years ago a car-maker launched a viral ad where people waiting to cross an intersection were entertained by a dancing figure on the red pedestrian traffic light.

Nobody likes to wait”, states the ad: “That makes traffic lights the most dangerous spots for pedestrians in the city” it warns. The commercial proceeds to show people dancing with the red-light figure, and then celebrates that the intervention led to 81% more people waiting at the red light.

Days after the ad was originally uploaded to YouTube, CityLab and other urban affairs websites were excitedly featuring it as an innovative urban intervention, boasting about the 81% increase in compliance, etc.

Here I will take a more critical look at it.


To justify the value of their undertaking, the ad makes three assumptions which are deeply ingrained into many people’s perceptions of pedestrian safety:

  1. That intersections are the most dangerous spots for pedestrians
  2. That this is so because pedestrians often cross when they are not supposed to, and
  3. That if pedestrians wait for their turn, they will be safe

So, how much truth is there to each of these assumptions? Let’s use Ontario’s Office of the Coroner (2012) Pedestrian Death Review to find out*.

According to this robust investigation, approximately 33% of pedestrian deaths in the province took place at intersections. This shows that, indeed, intersections are among the most dangerous places for pedestrians**. Assumption 1 is sound.

What about assumptions 2 and 3? A close examination of all fatal collisions that took place within intersections suggests there is little truth to them. Only 36% of all intersection deaths were due to pedestrians crossing against their signal, while 64% involved pedestrians who were struck while crossing when they were supposed to. This occurred when vehicles were turning right (21%), left (21%), going straight through intersections (12%) and at pedestrian crosswalks (9%) (Ontario’s Office of the Coroner, 2012).

In other words, two thirds of pedestrian fatalities at intersections were due to drivers’ failure to yield when they were supposed to. Indeed,  Nobody likes to wait.

This alarming figure raises a few questions: Should pedestrians follow the rules even if there is no certainty that motorists will abide? Is it sometimes safer to cross against one’s signal (say, if there are no cars in sight)? Could infrastructure provide more protection for pedestrians in the event that drivers fail to yield?

Answering these questions necessarily requires that we dispose of our ‘common sense’ thinking about pedestrian safety (and pedestrian mobility in general).


“Could infrastructure provide more protection for pedestrians in the event that drivers fail to yield?” Everything in this particular intersection encourages drivers to speed through.

Here’s one last shocking take-away from the Coroner’s report: pedestrians in Ontario are more likely to get killed by a motor vehicle while standing on the sidewalk or shoulder of the road than by crossing against their signal. That’s right, 14% of all deaths involved motorists losing control of their vehicles and mauling pedestrians on sidewalks and shoulders, while only 12% of all deaths could be attributed to pedestrians disobeying a red light. Turns out dancing on the sidewalk waiting for your turn might not be so safe after all.

Final thoughts:

An appropriate reading of the figures outlined in this piece does not suggest that crossing against one’s signal is safe (or safer). What it does suggest is that the assumption that pedestrians are generally struck and killed due to their own reckless behaviour is overwhelmingly refuted by data – at least in the Canadian province of Ontario.

Bonus Question:

Is it a coincidence that the ad  featured in this article was made by a car manufacturer?


*Note that, in basing our analysis on this document, we are choosing to consider ‘death’ as the chief variable in determining safety, rather than the prospect of a minor collision. It would be interesting to see how similar data from other world regions align.

** The second most likely location for a pedestrian death is ‘mid-block’ (31%): anywhere on a road where there is no controlled crossing. It is difficult to determine what this means exactly in terms of individual responsibility given the long stretches of road without controlled crossings in North American cities.