Thank you.
I know you don't want me to go into Lansdowne, and I won't get into the details. I'm not going to name names or that sort of thing, but my generic criticism was simply that there were too many moving parts. There were too many different businesses in the package, with different business risk horizons and different demands. There was a parking garage. There was a rink, a football stadium, and a shopping centre, and they were trying to put it all into one P3.
I'm trying to now answer your question. I'm referring to the Vining and Boardman studies on this. The most successful P3s are what I would call, for want of a better term, “single object”. There aren't multiple businesses in the one P3. It's a hospital, or it's a CSIS building, or it's a bridge to P.E.I. It doesn't involve a hotel sitting on the bridge, with a hockey arena.
What I'm getting at is, the more complex they are, in the sense of multiple businesses or strategic business units in the P3, the more cumbersome, the more difficult to manage, and more likely to fail they are.
To answer your question even more concretely: it seems that the most successful ones are those where the object is very clear—a highway, an airport building, a bridge, a hospital. First off, we've had a lot of experience in building these things in the past, so the risks are easier to evaluate and estimate. Of course, the private sector has had enormous experience in years and years of building these.
The role of government is to do a lot of due diligence up front with the private sector, to specify those risks, and to try to anticipate the unforeseen problems. We are going to need to have more P3s to address the infrastructure deficit, because as the population ages and there are these greater pressures on budgets, which you know about, there is going to be this need to find innovative ways to finance and address it.
Parallel to that, I hope that we move towards more tolls, where it's possible or feasible to have toll roads, toll highways, and user-pay policies on those infrastructural assets that are—quote—private goods; that is to say, they're divisible and you can exclude people from using them if they don't pay the toll.