I've seen transit projects with hundreds of millions budget fell apart with nothing to show at the end, and have seen a transit project that doubled the cost and tripled the schedule to get to completion, and that was a good project.
I've seen transit projects with hundreds of millions budget fell apart with nothing to show at the end, and have seen a transit project that doubled the cost and tripled the schedule to get to completion, and that was a good project.
- Unit price contracts (much easier to track growth in cost, helps to identify and mitigate issues earlier)
- Vendor performance management (Contractors and consultants are given a score after project completion, shapes how bids are evaluated on future projects)
- Qualified bidders (contractors and consultants are evaluated beforehand, only allowed to bid on projects if they've done something of a similar scope/content before - ties in with the point above, smaller projects will allow bidding by contractors with less history/lower scores)
The system isn't perfect (since it relies on vendor history, it only works well with projects of scope and type that they do regularly - i.e. roads and sewers mainly), but it's a lot better than a simplistic "lowest bid wins"
The way a private company would solve the problem would be hire a bunch of experts and in-house a lot of the technical design and procurement work.