Calgary ran more than 17,000 intersections through an algorithm and produced a ranked list of where pedestrians are most likely to be hit. Then it recommended upgrades at 26 of them.
The scoring model weighs traffic volume, pedestrian volume, roadway geometry and posted speed, and it spits out a number between zero and 100 for every crossing in the city. The highest score in Calgary is 99.02, at Martindale Boulevard and Martindale Gate in the northeast. Second is 98.54, at Falshire Drive and Falton Drive, also northeast. The lowest score inside the top 50 still comes in at 93.55, which is the part of this that deserves a second look.
Fifty problems and twenty six answers
A spread from 99.02 down to 93.55 across 50 intersections is not a spread. It is a cluster. Statistically, the fiftieth worst crossing in Calgary is barely safer than the worst one, which means the line drawn at 26 is not a line the data drew. It is a line the budget drew.
The estimated cost of upgrading those 26 sites is about $6 million. Design work is possible in 2027 and construction is possible in 2028, which in municipal language means neither is committed. So the city has produced an excellent piece of analysis and attached to it a work plan that touches half the identified sites, no earlier than two years from now.
Meanwhile the human numbers sit there. Fifteen people died on Calgary streets last year. Six have died so far this year, alongside 283 collisions involving pedestrian injuries. Those are not abstractions to anyone who lives near the top of that list.
The northeast keeps showing up
The two worst scored crossings in Calgary are both in the northeast, in communities where more people walk, more people take transit and fewer households run two vehicles. That is not a surprise to anybody who has stood at Martindale in February waiting for a gap in traffic.
It is worth saying plainly that the model itself is a good piece of municipal work. Ranking every intersection in the city on a consistent basis is exactly what a data driven administration should be doing, and Ward 5 has been publicly supportive of making decisions this way rather than by whichever community association files the loudest complaint. Scoring by measurement instead of by volume of email is progress.
The gap is not in the analysis. It is in what happens after the analysis lands on a desk.
Six million dollars in a city budget
Set $6 million against the scale of what Calgary spends every year and the figure stops looking like a constraint and starts looking like a choice. This is not a transformative capital program. It is signal timing, curb extensions, lighting and crossing treatments at two dozen corners.
Taxpayers are entitled to ask why the list stops at 26 when the data does not. If the honest answer is that the other 24 sites cost more per fix, publish that. If the answer is that the capital envelope was set before the ranking existed, publish that too. Calgarians handle real numbers fine. What erodes trust is a study that identifies 50 dangerous crossings and a plan that addresses half of them without explaining the cut line.
This is the same pattern that shows up when Calgary builders wait half a year for a decision. The city is not short of capable analysis. It is short of the follow through that turns analysis into concrete, and the delay is where the public cost accumulates. The province’s flood infrastructure spending proved the opposite point this summer, that money committed early shows up as damage avoided later.
Publish the whole list and the whole timeline
Three things would fix the credibility problem without spending another dollar. Release the full ranking rather than the top 50, so residents can look up their own corner. Publish the cost estimate for each of the 24 sites left out. And attach a firm year to design and construction instead of the word possible.
Calgary built a tool that tells it exactly where the risk is. Having gone to that trouble, the city owes residents an explanation of why the answer stops at 26.
Should Calgary fund all 50 flagged intersections, or defend the 26 it picked?
