As someone who has studied traffic engineering in school and works in road design, I'd be very curious what studies these were.

Look into it, there’s a heavy increase in collisions where cameras are present.

Only place I've seen this data was as an example in school of what not to do - several states had low yellow times (1-2s shorter than Ontario's), and added red light cameras at large, wide intersections that took longer to cross than the yellow timer, meaning if you entered on a green you could theoretically get hit with a red light violation. But those studies were late 90s, early 00s.

Every piece of data I've seen shows either a reduction in speed (even post camera removal), or minimal change after removal.

Note that studies need to reflect current state cameras in Ontario - only allowed to be used in school zones, and need to have signage present indicating their use. They're not used specifically at intersections.

Additionally, the fees for traffic cameras go back to road redesign budget, which is used (on the projects I've worked on) to provide traffic calming measures like narrower lanes, AT facilities, etc. Cameras should be a stop gap measure, and are vastly preferable to an increase in the polices budget to have increased traffic enforcement presence.