JeffDowntown
Senior Member
- Joined
- May 28, 2007
- Messages
- 4,558
- Reaction score
- 3,189
Re: North-South Rail Link
But when you are truly "guestimating" without any real hard data, you also need to provide error bands -- which is never done with this type of social sciences model.
Garbage in, garbage out. You can prove anything you want to prove with the "right" estimates.
Careful, sure, but you when you have the tools available to make a good model you need to go ahead and try. Everything can be quantified in some way or another.
I don't think people would be so gun shy about these multi-billion dollar potential boondoggles if the math really checked and could be checked. But so far our experience the last 20 to 40 years has been that projects are low balled by about 1/3 of their eventual real costs and then the economic benefits aren't really measured. So maybe people just won't believe the numbers because they have been used to obfuscate for so long, but if we don't get back to trying to get to real models then nobody will make big investments.
And all the arguments about demand management also apply to mass transit. If you don't build it, they don't come. You are already talking about a highly developed part of the city where adding a N-S link probably won't attract new development. This isn't like building a new station on the line where you have to justify with ridership numbers and capital costs of the station alone.
We are talking about diffuse system wide effects with improvements to travel times between specific destinations and enabling other improvements and efficiencies. That's why I say quantify it. Run some numbers, publish the results. Heck if it is a software model or even spreadsheet model then publish the source code so it can be verified. So far I've seen a few compelling bullet points and a lot of things which are nice to haves.
If the numbers add up, then great let's build a tunnel.
But when you are truly "guestimating" without any real hard data, you also need to provide error bands -- which is never done with this type of social sciences model.
Garbage in, garbage out. You can prove anything you want to prove with the "right" estimates.