I spoke at Executive Committee today and one of my stronger points was that ETS was allegedly making a profit off core routes, which meant that core neighbourhoods were subsidizing sprawl with their user fees. Unfortunately, it looks like that was a bit off base due to a miscommunication in our Mayor's blog post (cost vs loss symmetry in wording).
Given the some recent data in the open data catalogue (which is from September of 2013), the annual ridership for the year ending Setpember 2013 was 86,305,040.
When we look at the 2013 operating budget, the total cost of operating transit was 438,875,000 (with 309,476,000 from city subsidies and 129,399,000 from fares.
Some back of the envelope calculations yield a cost-per-trip of about $5.09, which after fares is ~$1.85, which means the $1.30 in Mayor Iveson's blog post is almost certainly the city's loss per ride, not the actual cost.
However, these numbers are based on ridership data that is estimated based on fares, simply dividing the total cost to implement the system by the number of "rides". Users with passes don't pay fares as such, and users with passes are the heavier uses (who are also more likely to use the higher-frequency, higher-capacity routes I talked about). I'm unconvinced that the actual cost per ride is as high as these numbers project, because unless we have a more accurate collection method (which I don't think we do until smart fares come in) the guesses seem haphazard at best.