I hadn't realised there were a short piece of tunnel on that line. If they've done PPP, they'll be paying through the nose for that piece of tunnel compared to had they done it separately ... what are the Eglinton tunnels - about $40 million per km?
Evergreen is a PPP (DBF, I think) with tunnelling included in the main contract, AFAIK.
Not to single any posters out, but I think there's a lot of confusion in the above conversation. Underground transit construction of any sort (bored tunnels, cut and cover tunnels, stations...) is more prone to cost overruns than, say, building a park. But regardless of whether you're building a subway or a park, you as an owner ask yourself the same question: is it better to ask the contractor to assume all risk and have them bake it into a higher fixed-price contract, or ask for a lower upfront cost with the understanding that you're covering all overruns. (This is a bit of an oversimplification, because there are a lot of different spots on the spectrum of who assumes how much risk, and even with a full risk transfer to the contractor there's ultimately the risk of their bankruptcy that means the owner gets stuck.)
When trying to make the decision to pick between one of the above approaches, how commonly those types of projects around the world see cost overruns is a bit of a distraction --- contractors darn well know this too, and their bidding reflects it. All things being equal, on a low-risk type of construction the risk premium that the market prices in on a fixed-price contract vs a cost-plus contract would be small; on a high-risk type of construction the delta is wider.
By way of analogy, let's say you're a renter. No matter where in the city you live you face a choice -- you can pay renter's insurance on your home's contents against theft, or have slightly more cash in your pocket but carry the risk you'll get burgled. Paying for insurance is not necessarily a "good deal" in crime-ridden areas and a "bad deal" in crime-free areas, because the insurance industry knows the risks and charges you accordingly.
The theoretical argument for PPPs comes down to the idea that by giving a single contractor enough latitude to design and build and oversee all the moving parts they gain enough confidence in their ability to manage risk that during the competitive process they lower the risk premium priced into their bid.
So, in the case of the Evergreen line tunnel, I don't understand the argument that it would automatically have been far cheaper if it had been split out and tendered separately outside the fixed-price envelope. I don't know if the recent Seattle/Bertha experience might be colouring perception a little here, but TBM-based tunnelling in well-known soil conditions is not especially risky, at least relatively to any other elements of underground transit construction.